Defining
Uncertainty in Earth System Based Process Models of Source, Reservoir and Seal
Facies
Markwick, P.J.1, P.J. Valdes2
(1) Petroleum Systems Evaluation Group, GETECH,
The use of Earth System models to predict
the past distribution and quality of source, reservoir and seal facies, has
gained considerable interest over the past 10 years. But, such models are
experiments that must be continually tested against observations, and so an
important part of any application of models must be a systematic, quantitative
assessment of uncertainty. This can then be added to the overall risking used
by explorationists.
Here, we present some results for source
facies predictions using a series of sensitivity experiments designed to
examine the response of predictions to changes in the model boundary
conditions, in this case atmospheric CO2. The full matrix of potential
uncertainties is extremely large, and includes not only sensitivity experiments
for changing boundary conditions (including plate tectonic and palaeogeographic
variability, atmospheric chemistry, surface vegetation, orbital parameters),
but also postulated intrinsic model issues (inter-computer processing variability,
missing (or equivocal) dynamic processes), and uncertainties associated with
our current understanding and representation of the processes responsible for
source, reservoir and seal depositional facies. We consider both marine and
terrestrial systems.
The results clearly indicate that
different variables respond in different ways to model variability, and this
then has similarly complicated response in terms of uncertainty. For example,
marine productivity systems based on gross ocean circulation patterns are
relatively robust, but changes in local geography (and especially bathymetry)
can modify this pattern and especially its magnitude and seasonality, which in
turn dictates how much of this productivity is then converted to export POC
(particulate organic carbon).
Earth Systems models provide a powerful
tool for explorationists, but only when applied through an understanding of
process and uncertainty.
AAPG Search and Discover Article #90063©2007 AAPG Annual Convention, Long Beach, California