Well-Driven Seismic
Processing
and
Reservoir Characterization
By
Stephen Patrick Morice1, Stefano Volterrani2, Tarek Nafie2, Ayman Shabrawi2
(1) WesternGeco, Gatwick Airport, United Kingdom (2) WesternGeco, Cairo, Egypt
Our “Well-Driven” approach to surface-seismic
processing
integrates borehole
data
throughout the
processing
sequence to achieve accurate depth images with
enhanced resolution, and constrained reservoir attributes from the seismic and
well
data
.
Borehole
data
are used to guide pre- and post-stack seismic
processing
testing by quantifying the match between migrated test volumes and the synthetic
seismogram or VSP corridor stack (e.g. Poggiagliolmi, 1998; Scott et al., 1999).
Attributes based on the correlation between well
data
and the seismic
data
at
the well location, and on characteristics of the extracted wavelet provide
objective criteria for
processing
parameter selection.
Intrinsic
processing
parameters are derived from analysis of VSP and log
data
. An Earth model of P- and S-wave velocities, densities, attenuation and VTI
anisotropy is built at the well location and extended over 2D or 3D. The model
is used to drive offset-dependent attenuation compensation, geometric spreading
corrections, demultiple operators, long-offset move-out corrections,
ray-trace-based muting, travel-time tables for pre-stack depth migration and AVO
analysis (Leaney et al., 2001).
Crucial to our method is appropriate conditioning of well
data
. Sonic and
density curves are edited using a multi-well regression scheme through
interactive, iterative calibration to VSP and surface-seismic
data
.
With the seismic
data
processing
sequence optimized to the borehole
data
, we
may proceed with greater confidence into seismic interpretation and the
derivation of reservoir attributes through calibration and classification of
seismic attributes with petrophysical and rock-physical borehole
data
. This
paper describes our borehole-integrated
processing
and reservoir
characterization methodology, showing examples from several recent projects.