Temporal biogeophysical signatures at hydrocarbon contaminated sites associated with long-term remediation efforts
Biogeophysical signatures of hydrocarbon contaminated sites provide ideal laboratories for investigating microbial-geophysical relationships as the excess organic carbon present at these sites stimulates microbial activity. As such geophysical investigations have documented characteristic changes associated with hydrocarbon biodegradation in both field and laboratory experiments. The conceptual model that results from almost a decade of studies from these environments is one in which over time, the geophysical signatures due to bio-physicochemical changes imparted on the aquifer by the microbial activity reach some maximum or minimum related to the availability of terminal electron acceptors, the organic carbon source concentration, and microbial activity. However, with continuous removal of the contaminant mass either by natural attenuation (e.g., intrinsic bioremediation) or engineered (bio) remediation, a decrease in the microbial activity is predicted to cause associated changes in the geophysical properties (i.e., geophysical signatures revert to original conditions). This paper will present the results of repeated geophysical investigations at a hydrocarbon contaminated site acquired over an eleven-year period documenting changes in geophysical signatures associated with removal of hydrocarbon mass in the contaminated zone. Initial investigations at the site showed that relative to background, the contaminated area was characterized by higher bulk electrical conductivity, positive SP anomaly, and attenuated GPR reflections. Over time, the contaminated zone bulk electrical conductivity had reverted to near background conditions, the positive SP anomaly became more negative, and the zone of attenuated GPR reflections showed increased signal strength. The removal of hydrocarbon mass in the vadose zone over the plume by a soil vapor extraction system decreased the level of biological activity and therefore the magnitude of the geophysical signatures. We conclude that the attenuation of microbial activity due to organic carbon mass reduction by natural or engineered (bio)remediation can be effectively imaged by temporal geophysical surveys.
Monitoring of a Shallow Gasoline Release using High Frequency Ground Penetrating Radar
This presentation reports the initial findings of our hydrogeophysical field experiment to evaluate the ability of high frequency (450 & 900 MHz) ground penetrating radar (GPR) to detect and monitor shallow light non- aqueous phase liquid (LNAPL) contamination. During August 2008, a mixture of 85.5% API gasoline, 10% ethanol and 4.5% MTBE was released into the unconfined sand aquifer at Canadian Forces Base Borden. The release is contained within a steel sheet piling cell that is placed into the underlying aquitard approximately 6 meters below. The water table was at a depth of 0.5 m at the time of the release and has varied between depths of 0.2 m and 0.7 m during monitoring. The initial GPR images of the LNAPL release zone reveal intense scattering of EM waves that obscure the underlying stratigraphic reflectors; the intensity of this scattering dissipated over the first month. This response indicates that the initial LNAPL emplacement resulted in substantial source zone heterogeneity that decreased over time. Only minor variations in GPR imaging were observed until the development of the frost zone, at which point the GPR signature of the LNAPL impacted zone dramatically changed. Two-way traveltime variations for an underlying stratigraphic reflection show a velocity "pull-up" under the LNAPL release during unfrozen conditions due to the displacement of water by the immiscible contaminant. Traveltime measurements changed with development of the frost zone, suggesting less freezing in the LNAPL zone.
Surface time-lapse electrical resistivity tomography (TLERT) monitoring of an SRS injection and associated biogeochemical processes, Oak Ridge National Laboratory, Tennessee USA
We present results from surface time-lapse electrical resistivity tomography (TLERT) data collected within a
uranium-contaminated unconfined aquifer underlying the Oak Ridge Field Research Center (ORFRC) located
at the Oak Ridge National Laboratory (ORNL) in Tennessee. As part of an Integrated Field Research
Challenge (IFRC) project supported by the DOE Environmental Remediation Sciences Program (ERSP),
bioreduction of U(VI) to U(IV) with ethanol as an electron donor has been tested during the last four years. Low
U concentration (below US EPA MCL of 0.03 mg/L) can be achieved by frequent injection of electron donor. To
reduce the costs and improve the sustainability for remediation and site maintenance, our IFRC team is
exploring the effectiveness of a slowly degrading substrate such as commercial emulsified vegetable oil
substrate (EVO) as alternative electron donor sources. Laboratory batch and flow-through column experiments
were carried out to investigate the sensitivity of various physical properties (e.g., electrical conductivity) to EVO
injection to test the applicability of geophysics as a monitoring tool at the field scale. Results revealed
increased electrical conductivity during both EVO injection and subsequent degradation of surfactant with an
overall increase in conductivity of ∼35%; thus, surface TLERT was selected as a monitoring tool to
supplement well fluid samples.
The field stimulation test began at Area 2 during early February 2009. Prior to the injection of the EVO,
preliminary characterization completed, including a geochemical survey of the ground water from ∼50
wells, microbial samples of groundwater and sediment collected from selected wells, and site hydrology
characterized by bromide tracer test and surface ERT methods. On February 9, 2009, diluted EVO solution
(20% concentration, 900 gal vol) was injected into three injection wells within 1.5 hours. Distribution of the
injected EVO and accompanying biogeochemical processes has been monitored since injection through
analysis of numerous well fluid samples and TLERT data from 2 profiles. Initial TLERT data were collected at
the 2 profiles over a two-week period at 12 different time steps. The surface profiles, situated parallel to and
perpendicular to the major flow direction (as delineated by tracer tests), are each 40 m long and consist of 52
electrodes spaced at 0.75 m. Initial analysis indicate good correlation between well fluid samples and TLERT
data and allow for improved extrapolation of well data to the field scale. Long-term monitoring is in place to
track the continuing hydrologic dynamics and reduction duration in this test area throughout Spring 2009.
A High Resolution Column Study of the Geoelectrical Profiles in a Gasoline Impacted Sand
A major objective of hydrogeophysics is the characterization of sites impacted by light non-aqueous phase liquid (LNAPL) contaminants. To obtain this goal, an understanding of the nature of the geoelectrical profiles at these sites is required. We have conducted a large column experiment to investigate the changes in the geoelectrical profiles due to the initial injection of gasoline and subsequent fluctuations of the water level. The high-resolution vertical geoelectrical profiles were obtained using closely spaced resistivity electrodes and time domain reflectometry (TDR) probes along the column. The geoelectrical measurements indicate as significant reduction in the capilliary fringe thickness during the initial gasoline injection. The first cycle raised and lowered the water level back to its original level. When the water level was established above its original level, the geoelectrical profiles clearly showed the presence of entrapped gasoline below the newly established water table. However, the profiles after this cycle was completed were nearly identical to the pre-cycle profiles, suggesting that these geoelectrical methods cannot differentiate between air and gasoline in the residual water saturation zone. The second cycle lowered and raised the water level back to its original level. In this case, evidence of the entrapped gasoline below the original water table were clearly seen in the geoelectrical profiles after the cycle is completed. A particularly interesting observation was made when the electrical conductivity profiles obtained from the resistivity electrodes and TDR probes were compared. While the profiles match in the saturated zone, they systematically diverge above the capillary fringe and reach a maximum difference in the residual saturation zone with the high frequency TDR values being greater than the low frequency electrode values. This divergence has significant implication for the inference of water saturation from electrical conductivity profiles
Investigating the Effects of Biofilm Development in Porous Media on Seismic Wave Propagation
Bioclogging, resulting from biofilm development is an important phenomenon that can cause significant changes in the porosity and permeability of subsurface systems with implications for fluid flow and contaminant transport. As such, a number of numerical models and simulations have been developed in an attempt to qualitatively forecast the effect of bioclogging on hydraulic properties. Limitations exist, however, with the application of these models as bioclogging processes are dynamic and quantitative information from the direct observation of biological growth and clogging is often unavailable. Here, we report on the results of a laboratory column experiment in which a minimally invasive acoustic wave imaging technique was used for the spatiotemporal characterization of biofilm development in porous media. Biofilm development was stimulated in silica sand-packed columns using a Pseudomonas aeruginosa PAO1 bacteria culture and acoustic (compressional) wave data were collected over a two-dimensional region for 29 days. In addition, complex conductivity measurements were collected to assess the progress of the stimulated microbial growth. The results from the biologically stimulated sample (nutrients and bacteria inocula) exhibited a high degree of spatial variation in the acoustic amplitude measurements. Portions of the biostimulated sample exhibited an increase in attenuation (up to 73%), while other portions showed a decrease in attenuation (~45%). The acoustic signals measured for the unstimulated sample (nutrients only), however, were relatively uniform over the 2D scan region. Environmental scanning electron microscope (ESEM) imaging of sand from the biostimulated column collected upon termination of the experiment verified the presence of biofilms on sand surfaces. ESEM imaging also revealed apparent qualitative differences in the structure and/or thickness of biofilm material between areas of variable acoustic wave amplitude. We infer from these observations that enhanced microbial growth and the presence of biofilms in the sand columns had a variable affect on the spatiotemporal elastic properties of porous media. Our results suggest that acoustic measurements may provide diagnostic semi-quantitative data for the validation of bioclogging models and numerical simulations.
Hydrogeophysical Characterization of the Vadose Zone of a Glaciofluvial Deposit as a Basis for Unsaturated Water Flow Modeling
Urban development exerts a growing pressure on soil and groundwater resources. Soil and groundwater protection in an urban context requires a better understanding of vadose zone hydrology at the scale of the urban site where the textural and sedimentary heterogeneity of the geological deposits may affect preferential water flow and contaminant transport. The goal of this study was to characterize sedimentary and flow heterogeneity of the vadose zone of a glaciofluvial deposit in the eastern part of Lyon (France) in which a stormwater infiltration basin was constructed. A hydrogeophysical approach was used based on the integration of a sedimentological analysis of the structural and textural organization of the deposit, a geophysical investigation using ground-penetrating radar and electrical resistivity tomography, and a characterization of the hydrodynamical properties of the deposit. Using these data, a three-dimensional model was built of a representative volume of the deposit underneath the infiltration basin. This model reproduces the deposit's heterogeneity at the lithofacies and hydrofacies scale. Water content measurements at three depths in the deposit, coupled with unsaturated flow modeling using HYDRUS, helped in understanding the hydrodynamics of the deposit during an infiltration event. Results show that the initial water saturation of the deposit has an influence on water flow at this scale and that flow heterogeneity is related to the presence of lithogenic heterogeneity. The hydrogeophysical approach used in this study has shown potential for the hydrostratigraphic characterization of heterogeneous sedimentary deposits.