Geological Aspects Of Model Construction for Well
Placement
Optimization in a Mixed Fluvio-Aeolian Reservoir
Caroline Y. Hern, Jonathan J. M. Lewis, Dirk Seifert, and
Nicholas C. T. Steel
The cost-effective development of small fields requires a thorough
description of the reservoir with a strong focus on the estimation of
uncertainty. in this example, 3-D reservoir models are used to define and risk
optimum well
trajectories in a complex, mixed fluvio-aeolian gas reservoir.
Emphasis is placed on the definition and quantification of those geological
factors that affect the Integrity of the hybrid stochastic-deterministic
reservoir model.
Sequential Indicator Simulation (SIS) models conditioned to well
data were
constructed for target zones. Construction of geologically realistic models
required identification of genetic units, an understanding of their spatial
distribution and assignation of appropriate dimensions. These objectives were
achieved by an integration of probe-permeametry and sedimentology data on
reservoir core, and the derivation of relevant genetic unit architectures from
analogous sequences at outcrop. Assignation of appropriate sizes depends on the
similarity between the outcrop analogue and the subsurface reservoir. Similarity
was assessed by comparison of facies
types
, their proportions, the vertical
association of facies,
types
of fluvio-aeolian interaction, the size of
depositional systems and gen ral depositional environment. Outcrop data were
compared to reservoir observations and sub-sampled to eliminate unrepresentative
elements.
The SIS models generated with this data were 'pin-cushioned' with over
100,000 numerical "wells" of varying azimuth, length, inclination and elevation.
Much improved decision support was provided through a statistical definition of
the 'optimum' well
trajectory.
AAPG Search and Discover Article #91019©1996 AAPG Convention and Exhibition 19-22 May 1996, San Diego, California