Ground-penetrating radar imaging of fluid flow through a discrete fracture

Abstract

Predicting groundwater flow and transport of contaminants in fractured rock is challenging due to the heterogeneity of hydraulic properties that are difficult to characterize using conventional hydraulic testing methods. Heterogeneity is often introduced by fracture aperture variability that creates preferential flow pathways also referred to as flow channeling. Ground-penetrating radar (GPR) has been used successfully for imaging fractures. This study investigates the polarization properties and capabilities of GPR signals, both amplitude and phase, for 3-D imaging of flow channeling in a discrete, subhorizontal fracture. Two separate field studies were conducted at the Altona Flat Rock test site in New York State. The first, conducted in 2010, used surface-based multi-polarization 3-D GPR to examine the effects of radar signal polarization for imaging a fresh-water saturated, millimeter scale subhorizontal fracture. Imaging of a horizontal reflection plane should be independent of radar signal polarization. However, amplitude variations as a function of wavefield orientation were observed along the subhorizontal fracture plane indicating that polarization effects are significant. Furthermore, it was shown that summation of two orthogonal parallel-polarized signals compensates adequately for the polarization effects and results in a more accurate image of the fracture. Therefore, for imaging of flow through a discrete fracture, multi-polarization GPR acquisition is necessary. The second investigation, conducted in 2011, utilized a multi-component, surface based GPR to monitor saline tracer flow through the same water-saturated fracture. The multi-component system allowed for simultaneous acquisition of orthogonal polarizations. The presence of saline tracer in the fracture resulted in an amplitude increase and phase decrease of the reflected GPR signal. Hydraulic dipole-flow tracer tests were used to generate flow between boreholes within the fracture of interest. A five-spot well configuration allowed control over the hydraulic gradient orientation. Various concentrations of saline tracer were utilized with hydraulic gradients oriented E-W and N-S, as well as along the natural gradient. GPR imaging of saline tracer results revealed a direct and narrow channelized flow path along the E-W orientation based on both amplitude and phase changes suggesting good well connectivity. The N-S dipole tests revealed greater tracer dispersion over a larger area suggesting poorer well connectivity. These results are supported by hydraulic tests conducted at the site. This work supports imaging tracer flow using GPR signal amplitude and presents, for the first time, imaging changes in flow channeling based on hydraulic gradient orientation and the use of GPR phase for imaging saline tracer distribution.