Removal of Overburden Channel Effects through Channel Velocity Modeling and Prestack Depth Migration for an Oil Field Offshore Abu-Dhabi

Abstract

To generate refined depth structure for a very gentle Cretaceous reservoir on an oil field offshore Abu-Dhabi, we conducted channel velocity modelling and prestack depth migration (PSDM). The target reservoir structure on a vintage prestack time migration (PSTM) contains pull-up and push-down artefacts, induced by a couple of overburden channel velocity fills. These artefacts hindered correct representation of gentle depth structure at the target. Detailed interpretation on the PSTM volume delineates overburden channels in three upper Cretaceous formations. One of them, Fiqa channel, generates small pull-up and the others, Halul and Lafan channels generate push-down below each formation, caused by higher and lower velocity channel fills, respectively. These artefacts overlapped at some locations. Analysis of synthetic PSTM stack from forward modelling reminded us that a low-velocity channel produces obvious push-down below the channel. The shape of push-down become smaller but wider with depth and offset. The resulted push-down in full-offset stack also changed with depth. A "true" channel velocity and implementation of PSDM eliminates push-down at all depths and offsets. Bearing in mind the phenomenon confirmed on the forward modelling, we first produced channel velocity model for the target field data. With the assumption that the depth structure at a layer just below channels is flat, time thickness of channel and the amount of push-down at the layer below the channel could determine the channel velocities. Then, we estimated three channel velocities by channel time thicknesses and the amount of pull-up/push-down at the top Mid-Cretaceous shale, Nahr Umr formation that located just below these channels. On the overlapped area, a linear inversion provided each channel velocity. An application of PSDM with the channel velocity model successfully mitigated the artefact at target depth structure. Introduction We sought refined reservoir structure through channel velocity modeling and PSDM at the target oil field, located about 60km off the coast of Abu Dhabi. The lower Cretaceous reservoir (Figure 1) structure consists of a couple of very gentle domal structure-highs, whose average dips are 1-degree and heights are less than 100ft. Since subtle difference of the reservoir depth made significant change in oil saturation and production rate, identification of such structure highs is critical for seeking enhanced productivity and oil recovery. Although vintage PSTM volume produced in 2004 showed good quality to provide more accurate reservoir structure and led to success on some wells, the time structure could contain as much as 10ms (60ft with 12,000ft/s) artefacts at the reservoir, caused by overburden channel velocity fills (Figure 2). These artefacts hindered correct representation of reservoir structure and gave rise to depth uncertainty. We realized that generation of precise reservoir structure necessitates discrimination between artefacts and true structural events. There are some publications that introduce solutions of these problems. Armstrong et. al., (2000) introduced a methodology of compensating artifacts using the amount of the pull-up and push-down changed with depth on post-migrated time volume. Fujimoto et. al., (2007) removed both pull-up/push-down and PSTM velocity artifacts through high-resolution tomography and PSDM.

Moveout inversion of wide‐azimuth data in the presence of velocity lenses

“…ment of the spatial extent of the amplitude and zerooffset time anomalies can be used to identify an LH layer on field data (Armstrong et al, 2001;Takanashi et al, 2008Takanashi et al, , 2009a.…”

Correction for the influence of velocity lenses on nonhyperbolic moveout inversion for VTI media

“…Nonhyperbolic moveout analysis is conventionally performed under the assumption that the overburden is laterally homogeneous on the scale of spreadlength. However, even gently dipping structures often contain velocity lenses such as channels and carbonate reefs with thickness between 50 m and 200 m and width smaller than the maximum source-receiver offset of a typical seismic survey (Armstrong et al, 2001;Fujimoto et al, 2007;Takanashi et al, 2008;Jenner, 2009; see Figure 1). For isotropic media, lateral heterogeneity of this type has been recognized as one of the sources of the difference between the moveout and true medium velocities (Al-Chalabi, 1979;Lynn Manuscript received by the Editor 29 July 2010; revised manuscript received 8 October 2010; published online 23 May 2011.…”

Moveout analysis of wide-azimuth data in the presence of lateral velocity variation

“…The pull-up time anomaly is observed for the bottom of the lens layer as well as for interfaces 3 and 4. The agreement of the spatial extent of the amplitude and zero-offset time anomalies can be used to identify an LH layer on field data (Armstrong et al, 2001;Takanashi et al, 2008Takanashi et al, , 2009a.…”