EFFECTS OF NAPL ENTRAPMENT DISTRIBUTION ON HYDRAULIC CONDUCTIVITY IN A RANDOM FIELD
|T.H. Illangasekare, H. Rajaram, and R. Compos, Department of Civil, Environmental and Architectural Engineering, University of Colorado, Boulder, CO 80309||
After a spill, a significant fraction of waste products that are in the form of nonaqueous phase fluid (NAPL) remains entrapped in the spread-zone of heterogeneous aquifers. These entrapped fluids act as long-term sources of contamination due to their low solubilities in water. Under rate limited conditions, the dissolution rates depend on the velocity of the flowing water in the soil.
Also, water velocity plays a critical role in determining the effectiveness of remediation schemes that use the water phase to deliver treating agents to the NAPL source. Examples of such schemes are enhanced dissolution using surfactants or co-solvents, thermal mobilization, and biotreatment. In this research, we have investigated the effects of NAPL entrapment on the spatial distribution of hydraulic conductivity. A set of experiments was conducted to simulate the flow and entrapment of a DNAPL in a randomly heterogenous aquifer.
Spills were conducted in two-dimensional tanks packed to represent realization of the same stochastic parameters of a random field. A dual-gamma system was used to determine the final entrapment distribution of the DNAPL. The measured saturation distributions were used to predict hydraulic conductivities using relative permeability functions.
These experiments show that due to unstable flow behavior (fingering), DNAPLs produce random saturation distributions in heterogeneous aquifers. This in turn, changes the distribution of hydraulic conductivity for the water phase significantly. It is our hypothesis that it is possible to predict the stochastic parameters of the final hydraulic distribution based on the parameters of the random hydraulic conductivity field prior to the spill.
In this paper, we present the experimental results and analysis of the results to test this hypothesis. The results of this study are of significant practical value as the developed methodologies will help in characterizing NAPL- contaminated sites using stochastic techniques rather than deterministic methods. Deterministic methods that have to depend on a large number of point-scale observations are not practical or cost effective in most DNAPL-contaminated sites.
Key words: NAPL waste, entrapment, effective parameters, remediation design, stochastic methods
Return to Main Table of Contents
Sub-Menu of Event Programs
Sub-Menu of Event Programs
Send comments on the Great Plains/Rocky Mountain HSRC web pages to: email@example.com;
comments or questions about this WWW server, to: firstname.lastname@example.org.