2006 Philadelphia Annual Meeting (22–25 October 2006)

Paper No. 5
Presentation Time: 3:05 PM

A NOVEL APPROACH FOR AQUIFER CHARACTERIZATION BASED ON DIRECT GROUNDWATER VELOCITY MEASUREMENTS AND GPR


SCHILLIG, Peter Curtis, DEVLIN, John Fredrick, MCGLASHAN, Michael Andrew and TSOFLIAS, Georgios P., Geology, University of Kansas, 1475 Jayhawk Blvd, Room 120, Lawrence, KS 66045-7613, schillig@ku.edu

A new technique in aquifer characterization is being employed in the relatively homogenous Borden aquifer to measure centimeter-scale heterogeneity and physical changes within a contaminant plume. In our experiment, a gasoline plume containing 10% ethanol was released into a controlled flow gate at CFB Borden, Ontario, in October 2004. Down gradient from the release, 20 innovative sensors capable of directly measuring point velocities in-situ were installed in multilevel fashion along the cross section of the gate. These point velocity probes (PVPs) provided spatially discrete velocity data at the centimeter-scale showing temporal velocity changes within the contaminant plume. Additionally, an array of GPR wells, surrounding a 2x4 m area enclosing the PVP fence, provided well pairs for high-resolution tomography, and a second, independent means of aquifer characterization through time. PVP velocities typically ranged from 6-15 cm/day, consistent with the calculated average linear velocity of about 10-11 cm/day for the Borden aquifer. GPR trends corresponded spatially with PVP trends; high groundwater velocities were associated with low EM wave velocities. When ORC was added up-gradient of the PVPs in October 2005, declines in groundwater velocity corresponded spatially to a zone of increased EM wave velocity. VOC data from multilevel samplers showed that these changes corresponded roughly to the area of greatest BTEX attenuation.

The combined use of PVPs with GPR indicated both the presence of small-scale heterogeneities in the aquifer, and the groundwater velocities within them. Furthermore, temporal trends measured by PVPs and GPR indicated that the two methods were complimentary and consistent. Data collected in August 2006 will assess the impacts of seasonal variations.