Sea‐bed diffractions and their impact on 4D seismic data

Abstract

A B S T R A C TA sea-bed reflection is used to estimate changes in water layer velocities between timelapse seismic surveys. Such corrections are crucial in order to obtain high-quality 4D seismic data. This might be a challenge in areas where rough sea-bed topography creates sea-bed diffractions that interfere with the sea-bed reflection, as in several of the northern fields in the Norwegian Sea. These diffractions and diffracted multiples are difficult to attenuate during data processing and become a source of noise in time-lapse data.In this work we study how cross-correlation analysis of time-lapse seismic data at a sea-bed reflection may be perturbed by the presence of diffractions. The seabed topography from the Norne field is used in 2D finite difference modelling to explain some of the observed variation in a water layer time-shift in field data. We find that a rough sea-bed and the diffracted energy it induces cause residual noise on the 4D data. The variation to the water layer time-shift increases with sea-bed complexity and is amplified by interaction with other sources of non-repeatability like water column variations, mis-positioning and strength of the ice scours creating the diffracted energy. Time-shift variations with mis-positioning and velocity changes between the surveys seem to best explain the observed variation in the time-shift for time-lapse seismic field data from Norne. I N T R O D U C T I O NThe importance of repeatability in time-lapse seismic data has been emphasized by the industry for many years. Landrø (1999) studied how repeatable a fixed-tool vertical seismic profiling (VSP) survey might be and how the repeatability changes with inaccuracies in source positioning. Harris and Adler (1999) investigated how the quality of the seismic data affects the accuracy of estimates of seismic variations in time-shifts, phase rotation or amplitude variation between time-lapse seismic data sets. Eiken et al. (2002) showed that repeated streamer positions were possible with new streamer technology that allowed for horizontal steering of the stream- * E-mail: sissel.grude@ntnu.no ers. Kragh and Christie (2002) examined the use of two repeatability metrics in assessing the similarity of two sets of repeat 2D lines acquired with a marine point receiver system. Eiken et al. (2003) showed that towing streamers with narrower than normal cross-line separation followed by spatial interpolation in processing can increase the repeatability. Calvert (2005) discussed various issues related to 4D reservoir monitoring and characterization. Hatchell, Wills and Landro (2007) showed that guided waves in a water layer can be used to accurately measure the water velocity changes at different times in a field. Lippard, Osdal and Riste (2010) tested to what degree 3D migration compensates the effect of mis-positioning in the final 4D image. Landrø, Digranes and Strønen (2001) combined traveltime and amplitude information from timelapse seismic data to map pressure and saturation changes for a non-compacting reserv...

Sensitivity analysis and application of time‐lapse full‐waveform inversion: synthetic testing and field data example from the North Sea, Norway

“…For a non‐flat water bottom, the multiples will be more complex due to focusing and defocusing effects and subsequently will be more difficult to model accurately. Furthermore, irregular seabed terrains cause seabed diffractions (Grude, Osdal, and Landrø ). Therefore, the advantage of having an FS may diminish when the water bottom is not flat.…”