Modeling Heterogeneities: A Quantitative Analysis of Two Slope Channel Systems: Laguna Amarga, Cerro Toro Formation, Chile and Canyon San Fernando, Rosario Formation, Mexico

Abstract

Outcrop-based studies offer the possibility of high-resolution characterization of both vertical and lateral variations within a channel belt. Complementing with shallow high resolution seismic data is likely the best way to reduce the key uncertainties that must be addressed when performing an evaluation of reservoirs. Three dimensional static model addressing heterogeneity distribution in two slope channel complex systems were performed using probabilistic techniques. The previously unmapped Cenomanian-Campanian Laguna Amarga/Lago Sarmiento Section of Cerro Toro Formation in Magallanes Basin, Chile, and the Maastrichtian San Fernando Channel System located within the San Quintin Sub-basin in Baja California, Mexico, offer a quasi-three-dimensional exposure that shows the spatial and temporal evolution of entire channel systems. Based on the geometries observed in one of the channel complex sets of these systems, conceptual 3D geologic models were created and compared using the observed facies distributions. Petrophysical data of analogues reservoirs with similar settings were used to populate the models and used as main inputs to the subsequent dynamic simulations. Simulation cases were defined to test the effect of well position, fluid type rock properties and internal heterogeneities on the recovery factor. The Laguna Amarga case shows slumps at the channel margins while San Fernando shows this facies in some cases capping the channels; an evaluation on the effect of these muddy facies on production was performed. In the same way thin-bedded turbidites with a net-to-gross less than 50% are shown to have an impact on production. Two main surfaces were extracted from the seismic data, while internal surfaces were created manually in order to use as reference during structural model phase. Object-based modelling constrained with pseudo-wells located in the different architectural elements, and a combination of geometrical parameters with general trends, were used to propagate the facies. Truncated Gaussian Simulations were also performed to test natural transitions of facies observed on the logs. Fluid flow simulations (depletion and waterflood simulation) have shown that upscaling in the vertical axis must be done at high resolution in order to preserve (< 1 meter) heterogeneities that may form barriers. Variations in permeability showed that these facies prevent water coning and have a big influence on compartmentalization within the channels. By modeling outcrop data it is possible to employ probabilistic methods to decrease the uncertainties associated with distribution of the main internal heterogeneities, and through flow-simulation studies to model their effect on production.

AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019