Universal Journal of Mathematics and Mathematical Sciences
Volume 5, Issue 2, Pages 149 - 158
(April 2014)
|
|
STOCHASTIC FLOW AND PARTICLE TRACKING MODELING FOR CONTAMINANT DETECTION IN SHALLOW, HETEROGENEOUS AQUIFERS
Evan K. Paleologos
|
Abstract: A high-resolution Monte Carlostochastic, coupled groundwater-flow-and-contaminant-transport model was developed to simulate the movement of chemicals into heterogeneous aquifers, resulting from random surface contamination locations. The objective of the study was to investigate the number of monitoring wells and the factors that affect sampling in order to maximize the probability to detect subsurface contamination. Our modeling addressed uncertainties in the heterogeneity and contaminant origination point(s) by describing groundwater flow in a stochastic framework with the hydraulic conductivity described as a log-normal, stationary, second-order random function, and the advection-dispersion equation simulated via the particle tracking, “random walk,” method.
Our work showed that the detection probability, of a monitoring network decreased significantly in heterogeneous, and strongly dispersive subsurface environments. The stipulated, by the U.S. and European regulations, number of three monitoring wells, even under optimal conditions, failed to detect subsurface contamination in more than 70% of cases. At least, 12 wells were needed if was to remain meaningful and even 20 wells, which in perfectly homogeneous media were shown to detect contaminants with certainty could drop to a detection performance of only one out of four cases in highly dispersive heterogeneous environments. In general, the U.S. and European regulatory requirements appeared to be inadequate with our study indicating a much larger number of wells needed for detection in commonly encountered geologic environments. |
Keywords and phrases: stochastic modeling, advection dispersion, random walk, groundwater. |
|
Number of Downloads: 244 | Number of Views: 633 |
|