P064 A New Approach to Warping for Quantitative Time–Lapse Characterisation P.R. Williamson* (Total Exploration & Production USA Inc) A.J. Cherrett (Total E&P) & P.A. Sexton (Elf Petroleum Nigeria Ltd) SUMMARY Time-lapse seismic images of producing reservoirs are often mutually displaced in time such that raw difference volumes may contain artefacts which are far from the reservoir. Standard correlation-based methods to align base and monitor surveys generate time-shifts which may sometimes be used interpretatively e.g. to assess compaction but in general have insufficient resolution and stability to be of interest as 4D reservoir attributes. Even after alignment the difference volumes while

Prestack depth migration is widely used as the best tool for imaging and velocity model QC and refinement, but is mostly applied isotropically, even when there is evidence of anisotropy. We outline a methodology of model-building and imaging for use in anisotropic contexts and apply it to a line of data from offshore West Africa dominated by massive shales. The results show that the anisotropy in the shale increases with depth, along with the velocity. The prestack migrated data are sufficiently sensitive to the anisotropy that a thin layer of sand showing little or no anisotropy can be detected, and they must be included in the model to get flat common image gathers everywhere. This suggests that the use of anisotropy as a lithology-discriminating attribute may be feasible on a depth scale of only a few wavelengths. In any case anisotropy determination is important for precise application of other lithoseismic methods such as AVO.

Abstract Accurate estimation of sub-surface structure is fundamental to exploration and reservoir characterisation. The majority of rocks found in the earth are anisotropic (i.e. the seismic velocity varies as a function of direction). This phenomenon can have a significant impact on the structure estimated from seismic data and yet is often neglected during seismic processing. In order to account for the anisotropy, a new model-based inversion scheme has been developed and adopted for the final development mapping of the Elgin-Franklin fields. For deep developments, such as Elgin / Franklin, the analytical depth conversion methods that currently dominate industry practice produce inaccurate structural interpretations, which significantly impact reserves assessment. The results of the model-based inversion, together with the corresponding uncertainty measures, have been incorporated into the development programme of the fields and the interpretation verified by subsequent drilling. Introduction Elgin/Franklin is a deep (5000m-6000m) Upper Jurassic, high temperature, high pressure gas condensate accumulation, located within the UK Central Graben of the North Sea (Blocks 22/30c and 29/5b). The Elgin field is a complex faulted anticlinal structure, while Franklin comprises a tilted fault block structure (figs. 1 and 2). The structures were first interpreted on a 1989 3D seismic survey which was of generally poor quality, especially in the NW and SE of Elgin, and the N and NE on Franklin. The deterioration of data quality below the base cretacious unconformity, strong anisotropy in the overburden and a high degree of faulting, especially in Elgin area complicated seismic imaging. In order to estimate volumes as accurately as possible, accurate and detailed structural imaging of the fields became a main issue: on Franklin, this constituted the improved definition of the main North-eastern bounding fault and on Elgin the delineation of the boundary and intra-field faults. It was decided, therefore, to acquire a new 3D seismic data set in 1996. The new seismic considerably improved the data quality and interpretability. The issue of depth positioning of horizons and fault delineation was addressed by building a 3D depth velocity model followed by 3D post-stack depth migration. Theoretical framework In-house software was used to build the depth velocity model 6,7. The stacking velocity inversion technique adopted is designed to replace the standard Dix-type inversion schemes that currently dominate industrial practice. The approximations used in these standard analytical methods of velocity estimation are well known. Velocity heterogeneity, finite offsets and anisotropy all contribute to a systematic mismatch between stacking velocities and the vertical root-mean square (RMS) velocity which are assumed to be identical. Instead, a model-based technique is used that allows many of these fundamental problems to be overcome successfully. The inversion is based around two stages of ray tracing (fig. 3). The first of these takes the picked horizons in migrated-time domain, backs-out the original time migration (a process known as demigration5), and then remigrates them into depth, using a hypothetical interval velocity field.

Summary Construction of velocity depth models should aim to integrate all available information in a consistent fashion to increase confidence in subsequent imaging and depthmapping operations. This often requires that the model be anisotropic. However data relevant to determining the anisotropy is usually acquired from limited regions at or around wells. We have extended our model-building software to include data from non-hyperbolic moveout analyses in the inversion, which improves this situation. We apply this method to a dataset from offshore West Africa. We attempt to check the quality of the model by performing prestack depth migration on a line which extends away from the wells and non-hyperbolic analysis line, and to calculate a model update from the image gathers. Strong multiples hinder a definitive assessment of the results.