A Geology-Based, Non-Seismic Attribute Method to Generate Facies, Lithology, and Petrophysical Parameters in the Chinook and Cascade Fields, Walker Ridge, Gulf of Mexico, USA

Abstract

The Petrobras America Inc. Cascade and Chinook fields are prototypical Gulf of Mexico deepwater (>8000 ft below sea level) Lower Tertiary Wilcox Trend oil fields with reservoirs starting at −25,000 ft subsea. The absence of production analogues for these high pressure, normal temperature fields with fine- to very-fine grained, low permeability, Paleogene sands required well production estimates to be based only on geologic computer models generated from analyzing well logs, pressures, fluids, conventional core, seismic 3D volumes, and structure maps. Although rock-physics studies indicated feasible lithology discrimination, very good well attribute versus quantitative seismic attribute correlations (such as acoustic impedance) have not been obtained, so far, for characterization of depositional packages. Consequently, an understanding of the geology was used to generate lithologic parameters.

Studies of the core have found that the depositional environment is a lower submarine fan with toe-of-slope to basin distal fine-grained turbidite deposits in channelized and non-channelized lobes. Analogs from Tanqua Karoo as well as medial-to-distal turbidite fields have shown that, where erosion is limited, layers with higher thicknesses are related to high sand content (high net-to-gross ratio), lower shaliness or clay volume in the sandstones packages, and higher porosity and permeability values. This principle was used as the basic guidance for the properties distribution (high net-to-gross ratio, volume of clay, and porosity) in Cascade and Chinook field models. Permeability distribution was controlled by porosity distribution according to the facies distribution and under volume of clay control. Either the thickness or quality of the sand were used to determine the turbidite facies (i.e., proximal or distal; amalgamated or heterolithic). Gross isochore maps from the seismic interpretation helped populate sandy areas horizontally, while well logs helped populate lithologies vertically. The geology-based method described in this paper was reasonably effective in generating geology computer models with parameters that were useful for the reservoir engineering analysis.

AAPG Datapages/Search and Discovery Article #90219 © 2015 GCAGS, Houston, Texas, September 20-22, 2015