Interpretation of depositional environments combined with field measurement of permeability for a portion of the Upper Cretaceous Straight Cliffs Formation near Escalante, Utah, provides new results for understanding and modeling facies-dependent permeability variations. Offshore, transition-lower shoreface, upper shoreface, and foreshore environments are interpreted for part of the John Henry Member on the basis of outcrop investigation. Using a newly designed drillhole minipermeameter probe, permeability was measured for two of the facies within this unit: lower-shoreface bioturbated sandstone and upper-shoreface cross-bedded sandstone. Approximately 500 permeability measurements at a sample spacing of 15 cm were made along four vertical profiles and three horizontal transects on a 6 m × 21 m outcrop.
Permeability ranges from 41 to 1,675 millidarcies (md) in the bioturbated sandstone facies, which is massive-bedded, moderately to well-sorted, and very fine- to fine-grained. In contrast, permeability ranges from 336 to 5,531 md in the moderately to moderately well sorted, fine- to medium-grained, cross-bedded sandstone facies. A high degree of variability in permeability of the cross-bedded facies is caused by small-scale variations in grain size and structure related to depositional processes. The geometric mean permeability in the bioturbated sandstone is 253 md, compared with 1,395 md in the cross-bedded facies.
While the facies-dependent differences in permeability (k) are apparent and related to depositional and biological processes, fractal-based statistical analysis of the horizontal ln(k) increments yields nearly identical results for the bioturbated facies and the cross-bedded facies, possibly suggesting an underlying statistical commonality in the formation of both facies. Increment distributions from both facies appear similar with peaking around the mean. Ln(k) increments from the smaller vertical data set appear similar also, but with a variance approximately 3.7 times larger than the horizontal value. Variance scaling analyses of horizontal and vertical data both yield a Hurst coefficient near 0.26, which is characteristic of negative spatial correlation of the increments. The methodology developed herein offers a potential high-resolution alternative to existing methods for understanding and characterizing subsurface properties.