Assessment of Scale on Permeability Estimates in Late Cretaceous Reservoirs, Denver Basin, Colorado

Christopher Rybowiak and Matthew Pranter

This study addresses how reservoir-scale permeability models vary depending on the scale of investigation of the input permeability values. A common practice in reservoir modeling is to directly use permeability measurements from core-plugs or probe permeametry in petrophysical modeling. The petrophysical models may have varying grid sizes but are often 5-7 orders of magnitude larger than scale of investigation (volume support) of the permeability measurement. This scale difference can produce unrealistic results in the petrophysical model and may not be representative of the reservoir heterogeneity. To explore this issue, two stratigraphic intervals, the Sussex (Terry) and Shannon (Hygiene) sandstone members of the Pierre Shale of the Denver Basin (within Wattenberg and Spindle fields), were selected for permeability analysis, near-wellbore modeling for effective properties, and 3-D reservoir-scale modeling. The stratigraphic units represent offshore bar sandstones and exhibit five common lithofacies. Permeability values by lithofacies (N=300 per core) from four cores were measured using probe permeametry and used as inputs for near-wellbore modeling to generate effective-permeability values using flow-based upscaling. The effective-permeability values exhibit a narrower distribution as compared to the original permeameter-scale values as well and show a shift in distribution towards lower permeabilities. Reservoir-scale, three-dimensional models [1 mi2 (1.6 km2)] of lithofacies, porosity, and permeability for the Shannon and Sussex were constructed for an area in the Spindle Field. Separate permeability models were generated using the original- and effective-permeability values by linking permeability values to porosity well logs. Using porosity and permeability cutoffs (>10% and >0.05md), the models were explored in terms of static connectivity of reservoir-quality facies and show the differences in connected volumes as a function of the input permeability values (original vs. effective). The models show differences in static "reservoir" connectivity as related to original- and effective-permeability that can be significant in terms of properly representing reservoir heterogeneity. The models illustrate the importance of scale of investigation when creating 3-D reservoir models of petrophysical properties as well as our upscaling method that shows this level of importance.

AAPG Search and Discovery Article #90156©2012 AAPG Rocky Mountain Section Meeting, Grand Junction, Colorado, 9-12 September 2012