--> Abstract: Permeability and Fluid Flow within Carbonate Parasequences, the Importance of Lateral Lithofacies Variation and Stacking Pattern--An Example from the U.S. Permian Basin, by P. M. Harris, D. J. Goggin, and C. W. Grant; #91012 (1992).

Datapages, Inc.Print this page

ABSTRACT: Permeability and Fluid Flow within Carbonate Parasequences, the Importance of Lateral Lithofacies Variation and Stacking Pattern--An Example from the U.S. Permian Basin

HARRIS, P. M., D. J. GOGGIN, and C. W. GRANT, Chevron Oil Field Research Company, La Habra, CA

Vertical and horizontal transects of permeability data were collected on core and outcrop from the upper San Andres Formation at Lawyer Canyon, Guadalupe Mountains, New Mexico, to assess the lithofacies and permeability variation in and between ramp-crest carbonate parasequences and what effect these variations have on waterflood flow simulations.

A typical parasequence (10 to 45 ft thick) consists of upward-shallowing facies with a basal dolomudstone (geometric mean permeability, permeability = 2.3 md), overlain by burrowed dolopackstone (permeability = 6.1 md) and bar-flank, ooid-peloid dolograinstone (permeability = 8.1 md). Thicker, better-developed parasequences have a capping (10 to 20 ft thick), bar-crest, ooid grainstone (permeability = 13.0 md). Therefore, an upward-increasing permeability trend occurs within each parasequence. Vertical correlation lengths of permeability, derived from semi-variogram analysis, support that parasequences may be the fundamental fluid-flow unit. The alternation through time between thick, grain-rich parasequences (fifth-order) and thin, mud-rich parasequences is thought to be the product f fourth-order sea-level fluctuations.

Three different permeability models were generated for a cross-sectional panel of three parasequences at the Lawyer Canyon site, and each model is input into a waterflood flow simulator to test the significance of the observed geologic variation: (1) a simple "layer-cake" permeability model distributed according to the vertical sequence of facies; (2) a geostatistical model that honors the vertical transect data, geologic markers, and the vertical and horizontal correlation analysis, but only minimally accounts for geologic variation; and (3) a geologic "truth" model incorporates k anisotropy and facies-specific horizontal variogram models to guide the geostatistics. These simulations show that the viscous-dominated flow behavior appears to be compartmentalized by individual paraseque ces and that the basal dolomudstones and poorly developed parasequences are potential barriers to fluid crossflow. In addition, sets of poorly developed parasequences (that is, those lacking well-developed bar-crest facies) have the highest potential for bypassed or remaining mobile oil. The three different models of permeability have similar overall oil recovery efficiencies; however, each model exhibits a drastically different processing rate, directly affecting the timing of recovery.

 

AAPG Search and Discovery Article #91012©1992 AAPG Annual Meeting, Calgary, Alberta, Canada, June 22-25, 1992 (2009)