Outcrop-Based 3-D Geomodel Analogue for Cretaceous Reservoir, Pecos River Canyon, Texas: Part 1: Facies and Rudist Mound Model

Janson, Xavier ¹; Kerans, Charles ²; Zahm, Christopher ¹; Playton, Ted ³
(1)Bureau of Economic Geology, University of Texas at Austin, Austin, TX. (2) Departement of Geological Sciences, University of Texas at Austin, Austin, TX. (3) Chevron Energy Technology Company, Houston, TX.

Complex carbonate mound architecture and especially the distribution of the generally coarser debris in three dimensions have a significant impact on reservoir-scale heterogeneities and flow paths, and therefore need to be realistically reproduced in 3D static models. Genetically linked facies distribution is hard to achieve using standard geostatistics because it violates the assumption of stationarity. This outcrop-based study compare commonly available geostatistical methods, including indicator Gaussian simulation, object modeling, and multipoint statistics to model complex Lower Cretaceous rudist buildups in 3D.

The study area is located in South Texas along the Pecos River Canyon that provides continuous exposure of the upper Albian strata for approximately 60 mi (~ 96 km). In the Lewis canyon area where one of the Albian high Frequency sequences is well exposed in multiple canyons, the transgressive systems tract is dominated by mud-rich facies that contain low-relief pancake-shaped buildups with a core colonizing community of chondrodontid clams and a capping facies of Radiolitid rudist rudstone and bafflestone. These mounds range from 1 to 5 m in thickness and between 30 and 1,000 m in diameter. By integrating detailed field mapping and ground-based Lidar of the outcrop, the different stratigraphic markers, mound core and mound flanks interval can be accurately mapped in 3D and used as conditioning data for the different geomodeling techniques.

The first model of the mound was done using indicator Gaussian simulation (IGS) with vertical proportions curves. The resulting simulation matched the vertical architecture observed in the outcrop but the circular mound shape in map view was not reproduced correctly. Next we used a combination of object modeling and IGS to model the mound core and debris using a puck-shape object that was 200 to 700 m wide and 3 to 7 m thick. Within each object, IGS constrained by proportion curves was used simulate the facies. The overall circular shape of the mound was well reproduced, as was the vertical stack but the lateral facies organization within each mound was not well reproduced. Finally, we used an MPS simulation with two conceptual training images that recreated the lateral and vertical facies organization and juxtaposition correctly but a significant effort is required to build the training images.

AAPG Search and Discovery Article #90135©2011 AAPG International Conference and Exhibition, Milan, Italy, 23-26 October 2011.