IMPORTANCE OF MATRIX DIFFUSION IN REMEDIATION PUMPING: SENSITIVITY ANALYSIS OF FRACTURE GEOMETRY AND MATRIX PARAMETERS

The current understanding of the effectiveness of pump and treat systems in fractured rock is hampered by limited data on fracture characteristics and networks in the subsurface. In order to best remediate these sites, hypothetical modeling can be used to identify key parameters that influence processes that dictate the effectiveness of the pump and treat systems in fractured rock. A hypothetical plume was created using Hydrogeosphere code in the center of a 10,000 meters by 10,000 meter grid with a node spacing of 100 meters. The east and west boundaries of the model are no flow boundaries. North and south boundaries are specified head boundaries to set up a background gradient. The plume was "remediated" with a withdrawal well under varying matrix, fracture and well parameters. Hydraulic conductivity, porosity, hydraulic gradient, dispersivity of the matrix, dispersivity of the fracture, aperture of the fracture and the network of fractures were varied. Sensitive parameters were identified by comparing the percent change in volume of the contaminant’s concentration at the end of the remediation phase. The sensitivity analysis of the matrix parameters showed porosity to be the most sensitive parameter followed by hydraulic conductivity, while dispersivity of the matrix was not sensitive. Decreasing fracture apertures to widths 0.04 mm or less had little to no effect on the concentrations. Increased connectivity of fractures lead to a rapid decrease in concentration followed by slower removal of contaminants trapped by matrix diffusion. Decreased fracture spacing improved total mass removal rate. Understanding the contaminant pattern in the matrix relative to the fracture will help determine an optimal location of remediation wells for different fracture networks.