Superfund Research Program
Geologic and Geophysical Characterization
Project Leader: Herbert Einstein
Grant Number: P42ES004675
Funding Period: 1995 - 2000
Three-dimensional representation of subsurface conditions is the basis for much of the assessment, modeling and prediction of subsurface contaminant transport and interaction in the Aberjona Valley. The preceding work on a computerized Kriging-Bayesian updating procedure was expanded to three dimensions. Because flow through rock fractures is potentially very important in the Aberjona Valley, research in this area was conducted involving the development of specific fracture system models. Point-wise exploration (boreholes) was substantially expanded with the geophysical techniques. Ground-penetrating radar (GPR) and electrical resistivity techniques were used to image subsurface electrical properties. These electrical properties were related to physical properties such as fluid content and porosity. Vertical seismic profiling (VSP) determined fracture presence and orientation in the bedrock. At Mystic Lakes, ground-penetrating radar identified sedimentary layers and disturbances in the sedimentation process.
Geophysical techniques, including complex electric resistivity monitoring, are also useful in detection of heavy metal contaminants in the groundwater. Since the measurement profiles can be observed at different times, it is be possible to observe changes with time, i.e. to actually measure the contaminant transport.
The Kriging-Bayesian updating procedure
Researchers expanded the procedure to three dimensions. In the fracture flow research, development continued on the combined geometric-mechanical model. This involved characterization of typical geologic mechanisms. In addition, a first set of algorithms for fractures due to folding structures was completed and implemented with field data. The U.S. Dept. of Energy sponsored this part of the research (involving an oil reservoir in West Texas). Another set of algorithms is being developed which represent faulting and are specifically aimed at modeling the fracture geometry in the bedrock underlying the Aberjona Valley.
To determine if GPR can extend point-source ground truth information into two and three dimensions, electrical conductivity values were measured at some of the same points where coring and cone penetrometer measurements were made. Investigators used these measurements to physically correlate electrical properties with sub-surface structure. Conductivities at coring locations were estimated independently by directly measuring the exhumed soil and comparing the values to the probe measurements.
Specially designed GPR surveys were carried out and electromagnetic (EM) velocities were estimated from constant moveout profiles (CMP) which were used to infer the depth to the reflective layers. A synthesis of the conductivity transition zones was then compared to the stratigraphic surfaces detected by the GPR surveys, as well as the penetrometer data. Finally, an empirical relationship between soil stratigraphy and attenuation of the radar signal at these isolated locations was assembled. Cone penetrometer measurements were made along two orthogonal lines against which GPR surveys along the same profiles could be compared. The results show a strong correlation between the probe data and the electrical measurements. In order to strengthen the correlations between the 2-D ground truthing data and the GPR images, extensive image processing codes were written. The subsequent images were then compared along the same profile lines and show excellent agreement with the penetrometer study.