Bachrach, Ran1, Nader Dutta1
(1) Schlumberger RS/DCS, Houston, TX
ABSTRACT: Seismic Reservoir Description and Uncertainty Estimation: Examples from Clastic Basins
Seismic reservoir description (SRD) effort is based on mapping seismic attributes such
as P impedance, shear impedance, and density into reservoir properties such as lithology,
saturation, and porosity. The problem is non-unique, and thus, uncertainty is associated
with the predictions. In many cases, geological knowledge can be used to reduce the
inherent non-uniqueness of the SRD problem. However, introducing such information in
quantitative manner can often be difficult.
In this presentation, we show two ways to incorporate geological interpretation and
modeling into seismic reservoir description. We use Bayesian lithology classification
techniques to derive lithology classes from rock physics, seismic inversion, and well-log
data analysis. We then show how structural interpretation derived from wave kinematics can
be combined with the seismic inversion derived from wave dynamics to generate prior models
that will be used within a Bayesian classification framework to enhance the resolution of
seismic reservoir description efforts. This approach can use geological modeling or
detailed seismic interpretation to derive the most likely lithofacies given seismic
response. We then show how we can extend this analysis in depth by accounting for
theoretical or empirical compaction curves to generate a depth-dependent classification
function. We show how accounting for this type of secondary information can reduce
uncertainty in seismic property description. We provide examples from two clastic basins
(deep water) where this methodology has been tested.
AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004