2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 2
Presentation Time: 8:25 AM

PREDICTING THE HETEROGENEOUS DISTRIBUTION OF AQUIFER PERMEABILITY THROUGH MAPPING LITHOFACIES


ALLEN-KING, Richelle M.1, DIVINE, Dana P.2, TAYLOR, Kathryn M.3, GAYLORD, David3 and ALLDREDGE, Rich4, (1)Geological Sciences, The Univ at Buffalo, Buffalo, NY 14260-3050, (2)Pacific Groundwater Group, 2377 Eastlake Ave E, Seattle, WA 98102, (3)Geology, Washington State Univ, Pullman, WA 99164-2812, (4)Statistics, Washington State Univ, Neill Hall 0413, Pullman, WA 99164-3144, richelle@buffalo.edu

Prediction of permeability (k) by way of lithofacies mapping techniques can be used to quantify subsurface heterogeneity. In this project, we test the hypothesis that hydrologic properties that control subsurface contaminant transport are correlated to lithofacies. The accuracy of using lithofacies to predict permeability was tested by comparing the statistical characterization of permeability (e.g. mean, variance, normality) by lithofacies determined for sediment samples for cores collected along two transects (parallel and transverse to groundwater flow) in the Borden aquifer (Ont., Canada). Lithofacies mapping identified ten lithofacies in the parallel direction, and nine of those in cores oriented transverse to the groundwater flow; the mean, variance, and shape of the permeability cumulative distribution function (CDF) were compared to those defined in the parallel orientation.

Samples from the cores along the section transverse to groundwater flow showed little variation in the permeability CDF from the parallel transect in six of the nine mapped lithofacies. Moderately- to well sorted, laminated and cross-stratified lithofacies returned predictable permeability measurements (e.g. similar mean, variance, CDF shape for samples from a particular lithofacies taken from each of the two core transects). The shapes of the CDFs are similar for most lithofacies, but three differ in their mean value and/or variance of k. The two lithofacies showing inconsistent mean k for two directions were the massive coarse grained and high-angle planar cross-stratified lithofacies. The massive fine-grained lithofacies showed an inconsistent variance of k for samples from the two transects. These three lithofacies are generally more poorly sorted than the other six lithofacies. For these lithofacies samples, it may be possible that relatively subtle differences in grain size and sorting cause the observed differences.

This comparison study shows that lithofacies can provide a framework to estimate the three-dimensional permeability field spatial distribution for a sedimentary aquifer. Such a finding presents opportunities to improve contaminant transport prediction and remediation efficacy.