Date of Award
1998
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Geology
First Advisor
P.J. Gerla
Abstract
Most methods of delineating capture zones for pump-and-treat remedial design and wellhead protection assume a two-dimensional, homogeneous aquifer. Aquifers, however, are three-dimensional and heterogeneous, thereby introducing uncertainty in capture zone analysis. This study used a Monte Carlo analysis of three sets of statistical parameters defining aquifer heterogeneity. Each set had a different variance for the mean natural log hydraulic conductivity, which varied among the sets by a factor of four. The ensemble means of the capture zones for each set were estimated from 10 randomly generated fields with 12 layers each which were superimposed on each other for a total of 120 realizations. Realizations of the hydraulic conductivity fields were generated using the fast Fourier transforms method (Gutjahr et al., 1996) and incorporated into a confined, 128 m long by 64 m wide grid that included sufficient vertical layers to maintain the vertical correlation length. Constant head and no-flow boundaries were established on the short and long edges, respectively. The U.S. Geological Survey MODFLOW code (MacDonald and Harbaugh, 1988) coupled with MODPATH (Pollack, 1989) were used to simulate the capture zones surrounding a pumping well within this simple flow system. Mapping the capture zones for many aquifer realizations with similar stochastic properties provided the data required to construct 1 %, 80%, and 99% quantile intervals. These figures begin to show how heterogeneity reduces the size of capture zones estimated for statistically homogeneous aquifers.
Recommended Citation
Jones, Jon Paul, "Determining capture zones for three-dimensional, heterogeneous, anisotropic aquifers by Monte Carlo analysis" (1998). Theses and Dissertations. 153.
https://commons.und.edu/theses/153