2004 Denver Annual Meeting (November 7–10, 2004)

Paper No. 11
Presentation Time: 4:25 PM


KINCAID, Todd, Hazlett-Kincaid, Inc, 505 Arlington Ave, Suite 203, Reno, NV 89509, HAZLETT, Timothy J., Hazlett-Kincaid, Inc, 6753 Thomasville Road, Suite 108-213, Tallahassee, FL 32312 and DAY, Kevin, Hazlett-Kincaid, Inc, 505 Arlington Ave, Reno, NV 89509, day@hazlett-kincaid.com

The fundamental weaknesses of flow and transport models lay in the site conceptualization and the articulation of that conceptualization in model space. Too often, key site complexities, such as three dimensional heterogeneities, subsurface structures, karst features, etc. are disregarded because they cannot be adequately incorporated in a conceptual model. As a result, critical complexities may be ignored despite the fact that, at most large and complex sites, there are data from numerous (tens or even hundreds) of boreholes, geotechnical probes and/or surveys available from which to develop a site-wide characterization. In many cases, the problem actually boils down to having more data than can be adequately processed and synthesized using standard methods.

This paper describes a probabilistic method for synthesizing disparate site lithologic and electrical conductivity data into a fully three-dimensional solids model of hydraulic conductivity distribution in a surficial aquifer, using the EarthVision™ (EV) modeling software and custom intermediate programs that interact with EV at various stages of model development. The resulting three-dimensional Geological Framework Model (GFM) was used at an industrial site to describe twenty-six uniquely identified USCS soil zones using more than 300 visual borehole logs. The model was then refined using electrical conductivity data where the borehole logs were sparse. The zones were then regrouped based on hydraulic conductivity values measured in the laboratory from Shelby Tube samples or those reported in the literature and then redefined to model the distribution of those hydraulic conductivity groups in three dimensions. Finally, the resulting three-dimensional hydraulic conductivity field was exported for upload into a finite-element numerical groundwater model.

Using this approach, we were able to fully capitalize on all available site data to describe the heterogeneity of the surficial aquifer in the flow modeling framework, thus providing for more realistic simulations of flow and transport at the site. In addition, significant insight into the processes controlling NAPL migration and localization was gleaned through the process of the GFM development.