AAPG ANNUAL CONFERENCE AND EXHIBITION
Making the Next Giant Leap in Geosciences
April 10-13, 2011, Houston, Texas, USA
Enhancing Subsurface Reservoir Models – An Integrated MPS Approach Using Outcrop Analogs, Modern Analogs, and Forward Stratigraphic Models
(1) Chevron Energy Technology Company, San Ramon, CA.
The Multiple Point Statistics (MPS) approach to reservoir modeling uses combinations of hard data constraints (well and seismic data) and geologic concepts (e.g., depositional models) to populate 3D grid space. Using a carbonate ramp example we combine information from outcrop and modern analogs with Forward Stratigraphic Modeling (FSM) to suggest strategies that can be incorporated into subsurface modeling workflows. Using this integrated approach, a ramp outcrop is successfully simulated wherein the sequence stratigraphic architecture, including stratigraphic partitioning of facies belts, ramp progradation and retrogradation, and changes in facies belt width are captured.
LIDAR and GPS data are integral for an MPS outcrop model to capture the stratigraphic framework and representation of complex stratal patterns, including: (1) environments of deposition (EODs) - inner, middle, and outer ramp; (2) lower frequency sequence architecture showing two cycles of aggradation to progradation; and (3) high frequency sequences displaying partitioning wherein HSTs are progradational, oolitic grainstones and TSTs are retrogradational to aggradational, peloidal mud-dominated wedges. Concepts from the outcrop are used in virtually every step of the MPS workflow, including Training Images, the Vertical Proportions Curve (VPC), and a Facies Probability Cube (FPC), which together capture juxtaposition relationships and 3D likelihoods of occurrence of ramp environments.
Modern carbonate analogs provide geobody attributes for Training Images and the VPC including widths of facies belts, configuration of facies patterns, and variation along strike. Flood tidal deltas in the Exumas of Great Bahama Bank occur within a >450 km2 linear belt with: a) delta lobes extending up to 8 km, averaging 6 km, onto the platform; b) large deposits being highly sinuous, more irregular in shape than smaller ones, and maintaining connectivity; and c) tidal channels averaging ~3 km in length with regular, but regionally variable, spacing. FSM provides conceptual attributes for Training Images and constraints on the VPC in terms of global proportions, and in this case a ‘Dionisos’ model simulates the outcrop and provides a quantitative template for interrogating the extent and distribution of potential reservoir and non-reservoir facies providing insight into dip length, thickness, and connectivity of the middle ramp EOD, HST and TST components, reservoir-prone and best reservoir facies.