| Authors |
Florent Lallier, Nancy Universite/INPL-INRA; Sophie Viseur and Jean
Borgomano, Laboratoire de Geologie des Systemes et Reservoirs Carbonates; and
Guillaume Caumon, ENS Geologie/INPL
|
| Preview |
This reference is for an abstract only. A full paper was not submitted
for this conference.
1) Introduction
In the static and dynamic workflow of carbonate reservoirs, stratigraphic
correlation of well data is one of the first and most influent steps. Indeed,
facies distribution and petrophysical properties mainly control flow simulation
and are often computed thanks to geostatistical methods, on grids based on
stratigraphic correlation and structural data interpreted from seismic data
(Borgomano [2008]). In reservoir uncertainty modeling approaches, a unique grid
is built, and uncertainties about layering geometry, facies distribution and
petrophysical properties are handled using multiple geostatistical simulations
(Charles et al, 2001). This article aims at assessing uncertainties due to
stratigraphic correlations by also generating several set of possible
stratigraphic well correlations. Several grids may then be built from these
results and used for facies and property modeling. The method presented here
generates automatically and stochastically sequence stratigraphic correlations
of carbonate ramp systems by hierarchically integrating multiple pieces of 3D
information as: (1) interpreted well data, (2) correlation lines extracted from
seismic, and (3) information obtained on analogs.
2) 3D Stochastic Stratigraphic Well Correlation method
To perform the correlation, we propose a multi-dimensional and stochastic
extension of the Dynamic Time Warping Algorithm (DTW, Myers et al., 1981) that
we call msDTW. The DTW algorithm provides a way to find the optimal alignment
between two time series [Myer et al., 1981]. This algorithm was used for the
correlation of two wells by Smith and Waterman [1980], Howell [1983], Waterman
and Raymond [1987], Griffiths and Bake [1990], Brown [1997] for example.
Lallier et al. [2009] presented an improvement of the DTW, making the
method stochastic and introducing a hierarchy to mimic the reasoning made by
sedimentologists when correlating well data (Fig. 1), and applied this method
to a carbonate ramp system.
|