H33D-1605
Does small variance of ln K (~0.3) represent less geologic heterogeneity? A recent lithofacies study at the Borden aquifer
The variance of ln K has been commonly used to indicate the degree of spatial variability of the hydraulic conductivity (K) distribution in a geologic system. For example, the sandy Borden aquifer with its small ln K variance (= 0.29, Sudicky, 1986), has been considered to be relatively homogeneous. Recently, we collected core samples and mapped lithofacies in the Borden aquifer as part of a collaborative research project to determine spatial cross-correlation between the properties of the aquifer that control reactive transport. A horizontal transect of seventy-two densely-spaced (0.25-1.0m) cores, including those previously collected by Allen-King et al. (2006), allowed us to determine the spatial distribution and lateral dimensions of lithofacies units, which were not clearly discernible in previous studies. Our lithofacies mapping shows the following sedimentary features: dipping strata, channel-like structures, gradational transition among facies units, erosional discontinuities in the low-K materials, and isolated occurrences of silt materials. Geostatistical analysis of the horizontal cross-transition probabilities at two hierarchical levels reveals the contribution of stratal architecture to the sample semivariogram. The lithofacies structures identified in the Borden aquifer imply a potentially important role in groundwater flow and transport processes even with this low variance of ln K. This suggests that the characterization of spatial variability based solely on statistical abstractions without consideration of geologic structure may neglect important and useful information.
H33D-1606
Pressure and Concentration Profiles in Lab-Scale Heterogeneous Porous Media
Experimental observations in lab-scale heterogeneous porous media are compared with MODFLOW simulations. Pressure and concentration profiles are measured experimentally in a square sandbox filled with glass beads. Heterogeneity is included by replacing one-quarter of the space, in either a corner or center, with a layer of contrasting bead sizes. Pressure is measured with open manometer tubes. Concentration of a chloride tracer is measured electrochemically with silver/ silver chloride electrodes buried in the system. Results show good agreement between numerical simulations and pressure observations, but less agreement for concentration profiles. It is speculated that the rules for mixing at flow transitions are not well described in MODFLOW.
H33D-1607
Numerical Modeling Study For Saltwater Intrusion On Sand Tank Experiment
Saltwater intrusion on freshwater in sand tank was investigated with a numerical model. Groundwater level changes due to a difference of density between freshwater and saltwater were simulated and compared with the laboratory physical model results. Two-dimensional porous media model was developed to compute steady state pressure head using a commercially available finite element based computation code. In this model, the Brinkman equations were adopted for momentum transport through pressure gradients in porous media and through viscous effects, which is not able to consider with Darcy's law. A 25 cm thickness of porous media reproduces the laboratory test aquifer, 192 cm long, 160 cm wide and 53 cm high, composed of well-sorted medium sand (D50 = 0.47 mm). For the model flow boundary condition, a stagnation pressure head boundary was set for both left side and right side of the domain (x-direction). The stagnation pressure boundary condition, which gives constant pressure head and zero lateral velocity (u = 0) at the boundary, reproduces a saltwater providing reservoir. Atmospheric pressure boundary and no-slip wall condition were assigned at the top and the bottom of the domain (y-direction), respectively. Freshwater was recharged at the center of the porous media filled with saltwater. After the groundwater flow reaches to an equilibrium state, the interface water level and the freshwater-lens was observed. Groundwater withdrawal scenarios were tested to access the effect of groundwater pumping on the saltwater intrusion. Scavenger-well, commonly used to control saltwater-wedge, was tested for development of the groundwater contamination relief. The numerical model computations showed positive agreement with the physical model measurements. With the verified numerical model, various pumping rates of groundwater withdrawal scenarios were simulated to investigate relationship between the pumping rate and the saltwater level change. This modeling study will provide valuable information to predict a groundwater system development induced by saltwater intrusion.
H33D-1608
Cross-Evaluation of Laboratory Permeability Tests by Dimensionless Analysis
Accurate estimations of transport parameters of engineered and natural barrier materials are indispensable for assessments of the safety of geological disposal of hazardous wastes including radioactive nuclear waste. In such projects, it is essential that the barrier materials be capable of retarding the migration of contaminants. When groundwater flow is dominant, the evaluation of hydraulic parameters, particularly the permeability and specific storage, is of fundamental importance in the safety assessment. The permeability and specific storage of synthetic and geological materials are measured using permeability tests; however, conventional test methods are routinely applied to barrier materials with relatively low-permeabilities with little knowledge about their applicabilities. In order to provide a theoretical basis for designing laboratory permeability tests, the applicabilites of currently available test methods were examined and cross-evaluated with emphases on the experimental time and the parameter sensitivity using a numerical approach. To allow for a general discussion, a series of examinations were conducted based on a dimensionless analysis using dimensionless parameters representing the hydraulic properties and dimensions of the specimen and the other experimental conditions. In the dimensionless-time scale, the experimental time and the parameter sensitivity of the individual test methods can be simulated and compared by varying a few dimensionless parameters representing the experimental conditions. From a series of examinations, it was found that the experimental time and the parameter sensitivity of the individual test methods can be ascertained and compared using a few dimensionless parameters. This research project has been conducted under the research contract with the Japan Nuclear Energy Safety Organization (JNES).
H33D-1609
Physical and Hydraulic Properties of Rock Specimens from Grimsel Test Site, Switzerland
The Grimsel Test Site (GTS) is located at an altitude of 1730 meters in the granite rock of Aar Massif in central Switzerland. It lies at a depth of around 450 meters beneath the surface and was established over 20 years ago by the National Cooperative for the Disposal of Radioactive Waste (Nagra) as a center for underground Research and Development (R&D) supporting a wide range of related research projects. Among many of them, an international cooperative project defined as the Long Term Diffusion (LTD) has been performing to obtain quantitative information on matrix diffusion in rock strata under in-situ conditions. A set of laboratory experiments is also planed to determine the physical, hydraulic and diffusive transport properties of rock specimens taken from the same test site. In this poster, we present the preliminary results of both physical and hydraulic properties of the rock specimens being tested, including the bulk density, porosity, specific surface area and pore distribution, microstructure, P and S wave velocities, electrical resistivity, air and water permeabilities. The results obtained in this study indicate that: 1) The porosity and wave velocities of Grimsel granite are relatively low compared to the average values of igneous rocks indicating that micro-cracks can potentially exist within grain minerals. 2) The air and water permeabilities obtained from the air and water permeability tests are consistent that illustrates the accuracy of both experiments. 3) Permeability is not a simple function of effective confining pressure. It is very sensitive to confining pressure especially at low pressure levels. Besides, the permeability is hysteretic depending on confining pressure, pore pressure and stress history. 4) Similar to the hydraulic properties, diffusive transport properties of rock specimens can also be significantly affected by the confining pressure conditions. Laboratory diffusion tests considering the effects of stress conditions are fundamentally necessary for predicting the process of in-situ diffusion test and for comparing the results to be obtained from the in situ tests in the near future. Acknowledgement: This research project has been conducted under the research contract with the Japan Nuclear Energy Safety Organization (JNES).
H33D-1610
Detection and Characterization of Hydraulically Active Fractures in a Carbonate Aquifer: Results from Geophysical and Hydrochemical Measurements Along a 260-m-deep Borehole in the Combioula Geothermal System, Western Swiss Alps
In mountainous regions, pronounced topography, complex geology, and highly permeable geological formations and/or deep-reaching faults and fractures largely govern the development of deep water circulations. In this context, fractured aquifers are of particular interest and the detection and hydraulic characterization of the fractures is a correspondingly important task. Here, in addition to traditional hydrogeological techniques, borehole geophysical measurements were conducted in a fractured deep carbonate aquifer located in the Combioula geothermal system in the western Swiss Alps. The objective of the study was to detect and characterize the hydraulically active fractures along a 260-m-deep borehole through the integration of a comprehensive suite of geophysical well-log data and hydrochemical analysis. Specifically, we wanted to relate the geophysical signals to the fracture network and to examine the sensitivity of the different methods to the flow of groundwater inside the fractures. Although a number of geophysical borehole logging techniques are known to be sensitive to the presence and physical character of fractures, most methods do not provide information with regard to their hydraulic activity. This problem is potentially alleviated through self-potential (SP) measurements, which exhibit a direct sensitivity to fluid flow. Our results illustrate the potential of SP logging for complementing other geophysical logging techniques for distinguishing between hydraulically active and non-active fractures as well as for determining the flow direction within the fractures. The results further indicate that the hydrochemical evidence proved to be critical for the interpretation of the SP measurements. This work may help to open the door for a quantitative interpretation of SP logs with regard to fracture permeability.
H33D-1611
In-situ Hydraulic Property Estimation With Minimum Chemical/Microbiological Disturbance for HC Modeling of Sedimentary Rocks
For the investigations with drill holes aiming at a modeling of groundwater flow and mass transport with chemical and microbiological interactions, the minimization of chemical disturbance caused by drilling and in-situ hydraulic test is necessary for the long-term monitoring of groundwater chemistry and microbiological conditions. The authors carried out 350m drilling with ultra-filtered and deoxygenated drilling water in stable sedimentary rock area, and applied fluid electric conductivity (FEC) logging, flowmeter logging to estimate hydraulic properties along the borehole. For the continuous monitoring, the multi-packer type pressure measuring and sampling system was installed to monitor hydraulic pressure and chemical/microbiological environments. By using the transient pressure change at the water sampling as a small pressure pulse, hydraulic properties of sampling intervals were estimated with numerical inversion. Chemical analyses of groundwater samples showed the discontinuous profile split by tuff layer in massive mudstone. From the results of FEC logging, the authors calculated relative transmissivities and hydraulic heads of representative inflow/outflow points, and these values showed good accordance with flowmeter logging and multi-packer pressure measurements. With the hydraulic property and hydraulic head profile obtained in these measurements, groundwater flow model that gave a reasonable accounts to the chemical profile could be constructed.
H33D-1612
Sinusoidal Testing of Multi-Layer Aquifers
Sinusoidal aquifer tests can be used to characterize multi-layer aquifers. Three sinusoidal aquifer tests were conducted at the Waste Isolation Pilot Plant near Carlsbad, New Mexico, USA. Aquifer tests were performed by pumping at a variable (periodic) rate and measuring the resulting amplitudes and phases (time delays) of water level fluctuations in nearby observation wells. Aquifer properties for the Upper and Lower Culebra Member are found by fitting a two-layer model to observation-well drawdown data. Aquifer properties are estimated by subdividing the vertical section into multiple layers to provide greater resolution within the aquifer. The problem is approached mathematically using a matrix formulation that employs eigenvalues and eigenvectors to evaluate the resulting Bessel function of complex matrix arguments. Applications using multiple equivalent properties include the need to interpret conditions when pumping is limited to partial penetration, or to conditions when multiple, non-adjacent intervals are pumped. The solution can also be used to interpret tests in aquifers where vertical variations in aquifer properties are expected.
H33D-1613
Spectral Analysis of Pumping Response in Observation Water Wells
During a constant rate discharge test a well is pumped for a period of time and the resulting hydrograph is used to estimate aquifer properties. In cases where the pumping effect is expected to be within the range of background noise it can be difficult to determine whether pumping has had an effect. We examine a sample of hydrographs from several wells in central Florida using an s-transform to identify periods of pumping response in overlying aquifers as compared to pumped aquifers. A series of synthetic drawdowns are superimposed on the hydrograph of a surficial aquifer monitor well where no pumping exists to examine the viability of this method. The spectral analysis results identify pumping response signals as well as natural signals inherent in groundwater data such as lunar cycles and daily fluctuations. An investigation of these natural signals in 14- years of hourly data is also presented to evaluate common groundwater frequencies ranging from daily to decadal. This investigation affirms the utility of spectral methods for identifying small scale responses embedded in aquifer test hydrographs.
H33D-1614
Using inverse modeling of aquifer tests to estimate hydraulic properties and heterogeneity of fractured sedimentary rocks at the former Naval Air Warfare Center, West Trenton, NJ
At the former Naval Air Warfare Center (NAWC), West Trenton, NJ, inverse modeling of aquifer tests conducted in dipping fractured mudstones of the Newark Basin is used to characterize the heterogeneity distribution of the rocks and to estimate their hydraulic properties. This flow modeling is a first step towards development of transport models to investigate the fate and remediation of extensive chlorinated solvent contamination at the site. Ground water underlying the NAWC flows through shallow weathered rocks and dipping sequences of mudstone beds. Water-level and hydraulic-test data suggest that weathered rocks and bedding plane fractures within fissile and laminated mudstones are the primary pathways for ground-water flow, and that less fractured massive mudstones act as flow barriers. To examine these hypotheses, a MODFLOW-2000 model of flow through the fractured rocks is developed and calibrated to short-term aquifer tests conducted by temporarily turning off the pumps in individual wells of a pump-and-treat system. Within the model domain, the area of interest is 400 m by 700 m in horizontal extent and 100 m deep. Horizontal model layers that are up to 20 m thick represent the weathered rocks, and inclined model layers that vary from about 1 to 14 m thick represent the dipping beds. A set of separate aquifer-test simulations using the same underlying model and parameters is calibrated to water-level rise observations from the tests. Results show that a calibrated model that explicitly represents only the major contrasts in rock properties is consistent with the geologic framework and produces a reasonable fit to the aquifer test data. Estimated hydraulic conductivity values are realistic in comparison to independent field data, and vary by orders of magnitude among the different rock types.
H33D-1615
Modeling of the Assiniboine Delta Aquifer (ADA) of Manitoba using the Groundwater Storage from GRACE
This paper investigates the use of GRACE (Gravity Recovery and Climate Experiment) moisture storages for modeling of the Assiniboine Delta Aquifer (ADA) of Manitoba, Canada. There are great promises from GRACE in capturing regional groundwater storages that are potentially used for modeling application. However, it is well known that these storages are difficult to measure over the scales needed for hydrological model applications. Therefore, prior to modeling the aquifer using GRACE moisture storages, the storages need to be downscaled in to regional groundwater storages using the measured groundwater head data available in the area. Previous studies in the ADA have shown that the downscaled moisture storage estimates compared favorably with the measured groundwater storage over the area. This study focuses on the modeling of the ADA aquifer using the downscaled GRACE moisture storages. These storages will be used to initialize, calibration and potentially steer the hydrologic simulation. The calibrated model then will be validated independently using the measured data. These validations will hopefully provide better explanations for the underlying reasons for the differences in model predictions and measurements. This will identify some of the key assumptions and uncertainties in predicting moisture storage, and so highlight topics for further discussion and research.
H33D-1616
An Updated Site Scale Saturated Zone Ground Water Transport Model for Yucca Mountain
The Yucca Mountain site scale saturated zone transport model has been revised to incorporate the updated flow model based on a hydrogeologic framework model using the latest lithology data, increased grid resolution that better resolves the geology within the model domain, updated sorption coefficient (Kd ) distributions for radionuclides of interest, and updated retardation factor distributions. The resulting numerical transport model is used for performance assessment predictions of radionuclide transport and to guide future data collection and modeling activities. The transport model results are validated by comparing the model transport pathways with those derived from geochemical data, and by comparing the transit times from the repository footprint to the compliance boundary at the accessible environment with those derived from 14C-based age estimates. The transport model includes the processes of advection, dispersion, fracture flow, matrix diffusion in fractured volcanic formations, sorption, and colloid-facilitated transport. The transport of sorbing radionuclides in the aqueous phase is modeled as a linear, equilibrium process using the Kd model. The colloid-facilitated transport of radionuclides is modeled using two approaches: the colloids with irreversibly embedded radionuclides undergo reversible filtration only, while the migration of radionuclides that reversibly sorb to colloids is modeled with modified values for sorption coefficients and matrix diffusion coefficients. The base case results predict a transport time of 810 years for the breakthrough of half of the mass of a nonreactive radionuclide originating at a point within the footprint of the repository to the compliance boundary of the accessible environment at a distance of ~18 km downstream. The transport time is quite sensitive to the specific discharge through the model, varying between 31 to 52840 years for a range of specific discharge multiplier values between 0.1 to 8.9. Other parameters of importance include radionuclide sorption coefficients onto rock surfaces, diffusion coefficient within the volcanic matrix, sorption coefficient onto colloids and colloid retardation factors. Model breakthrough curves for various radionuclides at the compliance boundary are presented along with their sensitivity to various parameters.
H33D-1617
Water-level fluctuation and its implication on the hydrologic cycle in the Gwangneung Supersite, Korea
For effective assessment and management of water resources, it is important to understand and quantify each component of the hydrologic cycle. A careful and detailed analysis of spatio-temporal variations in water levels in aquifers could reveal useful information on the groundwater system. This study is objected to understand the reasons and mechanisms of fluctuations. As a part of an interdisciplinary research project, HydroKorea, to ascertain the water cycle quantitatively, water levels have been monitored from shallow monitoring wells(G1, G4) with less than 1-m in depths and a deep well enclosing three monitoring wells (ft1, ft2 and ft3 screened at depths of 102m, 45m and 6m below ground surface, respectively). Monitoring wells are located in the Gwangneung Supersite, Korea. Water levels used in this study were monitored by 10-min interval from February to May in 2007. Water levels compensated for air pressure were analyzed using a Fast Fourier Transform(FFT) technique for power spectral analysis. Results show periodic variations in 11.38, 12.19, 21.33, 24.38, and 28.44 hours, indicating strong influence of diurnal and semidiurnal tidal components. The diurnal components of the water levels from G1 and G4 in summer had greater power than those in winter, implying that the water table is affected by not only earth tides but evapotranspiration. However, those of the water levels from ft1, ft2 and ft3 do not show seasonal characteristics indicating that evapotranspiration has no influence in water levels of deep monitoring wells.
H33D-1618
Temporal change of groundwater temperature by a storm event
Possible seawater intrusion into the coastal unconfined aquifer caused by the passage of a typhoon was observed from the time-series data of water table, groundwater temperature and barometric pressure at the Niijima Island, Japan. We carried out continuous measurements of these data from 12:00 on June 16th to 16:00 on July 20th, 2007. A typhoon passed through the area at around 13:00 on July 15th. We observed significant drop of barometric pressure, rise of water table and increase of groundwater temperature by this event. To analyze the cause of this transient phenomena, we first estimated hydraulic diffusivity of the aquifer from barometric and tidal responses of water table before the passage of the typhoon. Then, height of sea level was estimated from barometric pressure and an astronomical tide table. Quantitative analysis using both estimated hydraulic diffusivity and estimated height of sea level suggested the sea water intrusion occurred by this event. Increase of groundwater temperature also supported this interpretation.
H33D-1619
A New Vulnerability Assessment Approach for Surface Water and Groundwater
As part of a state-wide effort to characterize the vulnerability of Mississippi's surface water and groundwater resources, we are developing a new set of vulnerability assessment tools. Unlike previous vulnerability assessment models (e.g., DRASTIC and WRASTIC), we separately consider three metrics for (1) intrinsic and (2) extrinsic sources of vulnerability and (3) consequences. Intrinsic vulnerability is defined as the vulnerability related to water resource characteristics (e.g., aquifer hydraulic properties, watershed topography, etc.), while extrinsic vulnerability is due to external features (e.g., number of abandoned wells, USTs, etc.). The consequence metric considers toxicity and location-specific human, environmental, and economic consequences. Each of the three metrics is normalized between 0 and 1, and the final vulnerability (risk) assessed as the product of vulnerability, susceptibility, and consequence. Intrinsic vulnerability is determined from the travel/residence time from a point on the land surface to surface water bodies or, in the case of groundwater, potential receptors. Unlike previous vulnerability assessment methods, our intrinsic vulnerability method focuses on assessment of relative travel/residence time and includes effects related to sorption, decay, and attenuation. Extrinsic vulnerability can be approached in two ways. In the first, a general extrinsic vulnerability is determined using a normalized indexing factor based on expert opinion and GIS-based density calculations for extrinsic sources of vulnerability (e.g., transportation corridors, industrial facilities, etc.). In the second, contaminant specific scenarios are considered based on local data. The consequence metric is determined using a toxicity-weighted measure of the potential impact on human, environmental, and economic assets. The vulnerability assessment methodology and GIS-based tools are being applied across the state at the hydrologic basin scale.
H33D-1620
Predicting Groundwater Recharge for the Okanagan Basin: A Little HELP From the Locals
Estimates of groundwater recharge are an essential component in flow models; however, recharge is highly uncertain and difficult to quantify for dry regions. Since flow models are often used in water management planning, acquiring spatially variable recharge estimates at a comparable scale can be challenging. Regional scale recharge estimates must reflect trends of local scale processes to be effective in planning. This study examined how results from a regional recharge model compared to recharge estimated in separate, local scale models from opposing ends of the Okanagan basin, in British Columbia, Canada. At nearly 8000~km2, the north-south trending basin has a climatic gradient with a warmer, wetter climate in the north and hotter, drier conditions in the south. The region has become a popular area for tourism, residence, and agriculture, particularly orchards and vineyards. With surface water sources close to fully allocated, the region is turning to groundwater as a means to support increasing socioeconomic growth. Basin wide (regional) recharge estimates in the valley bottom are a critical step in managing groundwater resources. Regional estimates of recharge were determined with the Hydrologic Evaluation of Landfill Performance (HELP) code by establishing common areas of soil texture, water table depth and three other hydrologic parameters. Results from the regional scale were compared with two independently derived, local scale estimates of groundwater recharge. For the south Okanagan, regional estimates were compared with results from a high-resolution integrated HELP and MODFLOW analysis; and in the north Okanagan, regional estimates were compared to results from a study utilizing the Richards equation based MIKE-SHE code. Comparison with these two models, calculated in areas at each end of the climatic gradient, provide confidence in developing a map of regional groundwater recharge. Preliminary results illustrate the applicability of HELP for predicting basin-wide recharge for areas with shallow slope (limited runoff), accurate depth to water table (from a water well database), and detailed descriptions of vadose zone lithology.
H33D-1621
Determining Seasonal Recharge to the Treasure Valley, ID using Dissolved Noble Gas Data
Recent municipal development and urban sprawl has placed strong stresses on regional groundwater systems, especially in western United States basin areas. Understanding of groundwater systems and fluxes is important for assessing the need for future water conservation practices. Quantifying these fluxes, however, has often been done by conceptual modeling or chemical mixing data, which supplies recharge origin, but not a well defined location. The use of noble gas concentrations to determine infiltration temperature and, comparing with the local lapse rate, elevation, has become very useful in these types of problems. Sampling of basin wells along with mountain front springs was done to determine the recharge source to the Treasure Valley Aquifer. Results have shown a distinct separation between mountain block infiltration and basin floor infiltration, evident in the spatial distribution of recharge temperature within the basin wells. This data confirmed previous work showing that a substantial portion of recharge to the northeast part of the Treasure Valley is the result of mountain block recharge and further delineates the spatial extent of influence by mountain block recharge into the basin. This data also highlights the importance of seasonally biased recharge both within the mountain block (spring snow melt) and at the basin floor (summer irrigation).
H33D-1622
Groundwater Mounding Beneath Stormwater Infiltration Basins
An accurate understanding of groundwater mound formation is important in the proper design of stormwater infiltration basins since these basins are often required to recharge a portion of pre-development infiltration volume. Mound formation due to localized recharge may reduce the infiltration rate of the basin and the ability of the soil to filter pollutants. The goal of this research was to understand groundwater mounding and the potential for contaminant transport resulting from recharge beneath stormwater infiltration basins. A 0.10 ha infiltration basin serving a 9.4 ha residential subdivision in Oconomowoc, Wisconsin was used in this study. Subsurface conditions included sand and gravel material and a groundwater table at 2.3 m below grade. Three storm events, 4.9 cm, 2.8 cm, and 4.3 cm, between August 2006 and April 2007 were modeled using the two-dimensional numerical model HYDRUS. The calibrated model was used to evaluate hypothetical basin operation scenarios for various basin sizes, soil types, ponding depths, and water table depths. The groundwater mound intersected the basin floor in most scenarios with loamy sand and sandy loam soils, an unsaturated thickness of 1.52 m, and a ponding depth of 0.61 m. No groundwater table response was observed with ponding depths less than 0.31 m with an unsaturated zone thickness of 6.09 m. The mound height was most sensitive to hydraulic conductivity and unsaturated zone thickness. A 7.6 cm sediment layer delayed the time to reach maximum mound height, but had a minimal effect on the magnitude of the mound. Mound heights increased as infiltration basin size increased.
H33D-1623
Residence Times and Pathway Analysis Using a Coupled Three-Dimensional Variably Saturated Groundwater Flow and Land Surface Model
The analysis of residence times and pathways of solutes in the subsurface is important in the characterization of biogeochemical processes and contaminant transport. Recent theoretical and experimental studies have shown power law (fractal) residence time distributions because of the fractal character of the topography. The theoretical studies focused on a steady-state analysis of the flow field utilizing an undulating water table boundary condition that follows the topography. However, processes of the vadose zone including root water uptake and evaporation have been neglected until this work. In this study, the three-dimensional variably saturated groundwater flow model ParFlow, coupled to a land surface model, is used to study the influence of processes of the vadose zone on pathways and residence time distributions. A small catchment is simulated incorporating topography; land and soil cover information; and one year of realistic atmospheric forcing. Transient Lagrangian transport simulations of a conservative tracer are performed, also including dispersion (i.e. heterogeneity), to develop spectral transforms of the arrival time distributions. The resulting power spectra show power law behavior over a wide range of scales. While the spectral scaling exponent (SSP) decreases with increasing heterogeneity, the influence of the vadose zone appears to cause an increase in the SSP. Additionally, the effect of explicit representation of subsurface heterogeneity and spatial model resolution on the scaling behavior is studied. This work was conducted under the auspices of the U. S. Department of Energy by the University of California, Lawrence Livermore National Laboratory (LLNL) under contract W-7405-Eng-48. This project was funded by the Laboratory Directed Research and Development Program at LLNL.
H33D-1624
Groundwater flow modeling using Lattice Boltzmann models
Lattice Boltzmann (LB) models are used to simulate flow in porous media at scales much larger than pore size. LB-based models for such macroscopic scale porous media flow simulations are an extension of standard LB models. There are at least two alternative approaches for implementing such models. In the first approach the local velocity is altered during the collision step by incorporating an external force, F, equivalent to the damping effect of solid particles in porous media. The porous media can be permeable or impermeable depending upon the external forcing term. The ability to simulate impermeable portions of a domain is validated using the Poiseuille equation for flow between parallel plates. Darcy's law is recovered for different permeability (k) and pressure gradients. Fractured porous media and karst domains with permeable matrix are simulated with this model. A sink term is introduced in LB model to simulate a pumping well and this model is further applied to solve steady state ground water well problems for confined aquifers. Directly solving the ground water flow equation with an LB model by exploiting its ability to solve diffusion equation is another strategy. The second order transient ground water flow equation is analogous to the diffusion equation and mass diffusivity is analogous to hydraulic diffusivity. This diffusion model is used to solve transient ground water problems. The simulated results show an accurate match with analytical solutions of the transient ground water flow equation.
H33D-1625
A Workflow Environment for Reactive Transport Modeling with Application to a Mixing- Controlled Precipitation Experiment
Advances in subsurface modeling techniques such as multi-scale methods, hybrid models, and inverse modeling, combined with petascale computing capabilities, will result in simulations that run over longer time scales, cover larger geographic regions, and model increasingly detailed physical processes. This will lead to significantly more data of increased complexity, creating challenges to already strained processes for parameterizing and running models, organizing and tracking data, and visualizing outputs. To support effective development and utilization of next-generation simulators, we are developing a process integration framework that combines and extends leading edge technologies for process automation, data and metadata management, and large-scale data visualization. Our process integration framework applies workflow techniques to integrate components for accessing and preparing inputs, running simulations, and analyzing results. Data management and provenance middleware enables sharing and community development of data sources and stores full information about data and processes. In the hands of modelers, experimentalists, and developers, the process integration framework will improve efficiency, accuracy and confidence in results, and broaden the array of theories available. In this poster (which will include a live computer demo of the workflow environment) we will present a prototype of the process integration framework, developed to address a selected benchmark problem. The prototype is being used to perform simulations of an intermediate-scale experiment in which a solid mineral is precipitated from the reaction of two mixing solutes. A range of possible experimental configurations are being explored to support design of a planned set of experiments incorporating heterogeneous media. The prototype provides a user interface to specify parameter ranges, runs the required simulations on a user specified machine, automatically manages the input and output files as well as a full description of the steps involved, and provides user interface to browse, visualize, and analyze the data. http://subsurface.pnl.gov/
H33D-1626
Sorption Impact of Chlorofluorocarbons to Carbonaceous Matter on Groundwater Age- Dating
The chlorofluorocarbons (CFC), CFC-11, CFC-12, and CFC-113, have been used as environmental tracers for groundwater age-dating. Most studies of CFC age-dating assume either conservative behavior (no sorption) or linear sorption with low KOC. However, sorption of other hydrophobic organic chemicals to thermally altered carbonaceous matter such as char, soot, and kerogen is exceptionally stronger than to bulk soil organic matter. Although CFCs have low water solubility, sorption studies of CFCs to various forms of carbonaceous matter (CM) have not previously been reported. This study focuses on determining the sorption of CFCs to two forms of CM, which is the first step in evaluating the effects of sorption on groundwater age-dating. Sorption experiments using batch techniques were performed with wood char as a representative of thermally altered CM forms and commercial humic acid as a representative of amorphous organic matter. Batch experiments using five different aqueous phase CFC concentrations ranging over four orders of magnitude were employed to develop sorption isotherms. The humic acid sorption reached equilibrium quickly while equilibration with wood char required about 100 times longer. Nonlinear sorption behavior to char was observed, but humic acid showed relatively linear sorption behavior. Comparison between wood char and humic acid KOC values indicates that sorption of CFCs to wood char is > 100 times stronger than sorption to humic acid at low dissolved concentrations. Considering the strong sorption to thermally altered CM, the retardation factor was greatly increased in comparison to the estimates from previous studies. These results suggested that CFCs can be significantly retarded even in low thermally altered CM contents. Therefore, this study shows that groundwater age-dating on the basis of CFC techniques may require consideration of retardation in aquifer environments containing thermally altered CM.
H33D-1627
Reactive Transport Modelling of Heterocyclic Hydrocarbons at a Former Gasworks Site
Highly mobile heterocyclic hydrocarbons constitute a persistent threat to groundwater at a former gasworks site in Southern Germany. This area is currently being used as a test site for novel subsurface investigation techniques, as well as for a site-specific enhanced natural attenuation (ENA) remediation approach. The subsurface investigation consisted of direct-push borings and monitoring well installations, tracer tests, and Integral Pumping Tests (IPTs) at multiple control planes, which were positioned at different distances downgradient of the source zone and perpendicular to the contaminant plume transport direction. The numerical inversion of groundwater concentration time series measured during the IPTs, in combination with a groundwater flow and transport model of the test site, leads to estimations of total mass flow rates of contaminants or other groundwater parameters relevant to NA, as well as of average concentrations and concentration distributions along the control planes. Following detailed investigations of the microbial and chemical conditions at the site, an in-situ ENA method involving a circulation well and the use of O2 as a terminal electron acceptor was devised to aid aerobic microbial degradation of the heterocyclic compounds in the otherwise anoxic surroundings. The aims of the study presented here were to verify the proposed conceptual model of interacting physical and biogeochemical processes and to assess the effectiveness of the in-situ remediation approach with the help of multi-component reactive transport modelling, focusing on the situations (i) prior to the implementation of the in- situ remediation scheme and (ii) following the operation of the circulation well with O2 injection. A finite-difference model was devised using the software PMWIN and PHT3D, utilizing the collected data to simulate the relevant processes in two and three dimensions, respectively. Instead of using interpolated point scale concentration measurements, the results of the IPT evaluations were used as input for the reactive transport model, thereby making it more reliable due to the integrating nature of the IPT method.
H33D-1628
Effect of Pore-scale Velocity on the Biodegradation of Contaminants during Transport in Porous Media
Column experiments were conducted to evaluate the effect of pore velocity on the extent of biodegradation of cis- dichloroethene (cis-DCE) during transport in porous media. The columns were filled with homogeneous glass beads and inoculated with the KB-1 culture (provided by SiREM, Guelph, Ontario, Canada), which is capable of complete dechlorination of perchloroethene to ethene. The columns were fed continuously with a synthetic groundwater containing a constant concentration of cis-DCE. Three different pore flow velocities (0.03, 0.08, and 0.51 m/day) were tested in duplicate, subjecting each column to a constant velocity for the entire experiment. Dechlorination of cis-DCE to vinyl chloride and ethene was monitored over time and space within the columns. Protein concentrations, also measured over time and space, were used to relate cell growth to biodegradation efficiency. At the end of the experiment, microbial DNA was harvested from the columns, and denaturing gradient gel electrophoresis (DGGE) was used to determine differences in the microbial communities that had developed in the columns subjected to different flow rates. The results show that the pore velocity has a strong influence on the microbial population and the degree of dechlorination. At high flow velocity (0.51 m/day), the degradation of cis-DCE to ethene was complete, and the organism capable of cis-DCE dechlorination ({Dehalococcoides sp.}) was present at the end of the experiment. In contrast, at medium and low flow velocities (0.08 and 0.03 m/day), incomplete dechlorination was observed with an absence or low concentration of {Dehalococcoides sp}. These results suggest that it is important for field-scale groundwater remediation to understand the interaction between physical and biological processes on the scale of single pores.
H33D-1629
Molecular Probes: A Tool for Studying Toxicity of VOCs to P.Putida F1
Volatile Organic Compounds (VOCs) are of great concern in ground water remediation, and are generally present in the form of NAPLs in subsurface environments. Among the various treatment technologies, in situ bioremediation is one of the most effective and low-cost treatment options. Many soil bacteria are reported to degrade these organic contaminants via metabolism (using them as a source of carbon to derive energy) or co- metabolism up to certain concentrations. However, larger concentrations of these contaminants are toxic to bacteria. Thus, in order to achieve successful bioremediation, it is important to determine the optimal concentrations of various contaminants that is beneficial for the activity and survival of degrading bacteria. The purpose of this study is to develop a novel method for toxicity analyses of VOC contaminants to the soil bacteria that degrade them. The present study is based on a two-color fluorescence assay of bacterial viability which facilitates actual counting of live and dead bacteria. Pseudomonas putida F1 cells were labeled with a LIVE/DEAD® BacLightTM bacterial viability kit (Invitrogen), which consists of a mixture of two dyes, SYTO 9 and propidium iodide, each with a different ability to penetrate healthy bacterial cells. Live cells stain green whereas propidium iodide (red dye) only penetrates cells with compromised membranes that are considered dead or dying. Stained cells were exposed to different concentrations of trichloroethylene (TCE) and toluene in sealed vials. Change in the concentrations of green and red cells were monitored over the time using fluorescence microscopy. UTHSCSA ImageTool software was used to count the live and dead cells in the images. It was observed that live (green) cell concentrations decreased and dead/damaged (red) cell concentrations increased over time when cells were exposed to TCE. No significant changes were observed in control experiments. Death rate constants calculated based on live cell disappearance and dead/damaged cell appearance were found to be approximately equal for TCE. Results will be presented in terms of dose response and death rate curves. Death rate constants and minimum inhibitory concentrations for survival of P. Putida F1 exposed to TCE and toluene will be compared.
H33D-1630
Mass Transfer from a NAPL Pool in the Presence of an Expanding, Mobile, Discontinuous Gas Phase: Experimental Study
The role of discontinuous, multi-component gas phases in groundwater systems is receiving increased attention. Recent research has demonstrated the significance of these gas phases with respect to non-aqueous phase liquid (NAPL) source zones where they have the potential to change our conceptual model for mass transfer at NAPL-contaminated sites. The spontaneous expansion of a discontinuous gas phase results from the presence of volatile NAPL compounds capable of generating gas-phase partial pressures that, combined with the partial pressures of other dissolved gases, exceeds the sum of hydrostatic and capillary pressures. These expanding gas phases overcome the capillary trapping forces, and are driven vertically over macroscopic length scales by buoyancy forces. This recently reported mechanism has the potential to increase the mass transfer from NAPL pools, and significantly change the dissolved NAPL concentration distribution, as compounds in a mobilized gas phase partition back to the aqueous phase well above the pool surface. Failure to consider this mechanism at sites where it plays an active role could result in the erroneous interpretation of aqueous concentration data and the calculation of inappropriate field-scale parameters based on a two-phase (NAPL-water) assessment alone. This work presents a laboratory investigation of this mechanism using small-scale (1.5 mL) no-flow reactors and an intermediate-scale (60 cm x 70 cm x 1 cm) flow cell. The small-scale reactors provided proof-of-concept information regarding the expected behavior for the expansion of a discontinuous gas phase in the presence of NAPL. The intermediate-scale flow cell was packed with silica sand and contained an emplaced NAPL pool. Multiple sampling ports were used to assess transient changes to the total mass flux and the vertical distribution of dissolved NAPL. In addition, light transmission imaging techniques were used to assess transient changes to the distribution of the discontinuous gas phase, including expansion, mobilization, and re-dissolution.
H33D-1631
Dense, Viscous Brine Transport in Porous Medium Systems
The use of high concentration brines in the subsurface has been studied for decades with a particular interest in storage of toxic and radioactive wastes. Brine has recently been incorporated into dense nonaqueous phase liquid (DNAPL) remediation strategies and evaluated at the laboratory and field scales. However, many open questions remain regarding the behavior and transport of such brines. We investigate the density and viscosity of a calcium bromide (CaBr2) brine as a function of mass fraction. We also evaluate dispersion characteristics using two one-dimensional columns packed with porous media that differed in microscale characteristics. Dispersion is observed for systems of varying velocity, density, and travel distance. Using a nonlinear transport model to describe fluid flow and species transport, a high-resolution finite volume discretization is applied to approximate a solution. The solution to the nonlinear transport model is then compared with experimental results.
H33D-1632
Composition and Properties of Coal Tar DNAPLs at Former Manufactured Gas Plants
Coal tar is a persistent source of groundwater contamination in the subsurface at many former manufactured gas plants (MGPs). Remediation of coal tar remains a significant environmental challenge due to its complex chemical composition, existence as a dense non-aqueous phase liquid (DNAPL), and tendency to alter the wettability of porous media systems. Changes in wettability are believed to occur due to the presence of ashphaltenes, a complex group of high molecular weight surface active compounds. Despite thousands of former MGPs in the US, a thorough investigation into the literature identified less than twenty sites from which coal tar DNAPL was analyzed to obtain data on chemical composition and/or physical properties. Currently, no study exists in the literature which considers multiple samples from a single site. In order to understand the potential universe of coal tar DNAPL and identify current data gaps, we compile all available literature data on coal tar DNAPL including: chemical composition, density, viscosity, and interfacial tension. We also conduct an analysis of multiple coal tar samples from a former MGP and evaluate changes in coal tar composition along the migration path of a DNAPL plume. Finally, we present a method for analyzing the asphaltene content of coal tar along with the results of this analysis.
H33D-1633
Modeling of Two-Fluid-Phase Flow in Porous Medium Using the Thermodynamically Constrained Averaging Theory Approach.
Building upon the general thermodynamically constrained averaging theory (TCAT) framework, we develop a model for two-fluid-phase flow. The model is formulated by averaging a general entropy balance inequality, conservation equations, and classical irreversible thermodynamics from the microscale to the macroscale to guide model closure. A closed model is derived under limiting assumptions, which includes among other distinguishing characteristics a connection between the pore scale physics and macroscale variables allowing for integration of experimental and theoretical results; and conservation equations for interface and common curve contributions to the system which are essential for modeling processes where capillary effects and inter- phase dynamics play an important role.
H33D-1634
Fractal and lacunarity analysis on soil structure and preferential flow using micro X-ray computing tomography
Quantification of soil structure and flow patterns is important to enhance our understanding of preferential flow phenomena in structured soil. Computing tomography (CT) provides a nondestructive means of observing soil pore structure and monitoring solute breakthrough in real-time. We investigated soil structure and solute transport dynamics in a large intact soil column. Five positions (at Ap1, Ap2, Bt, and two boundaries between the horizons) were scanned with a resolution of 78.1µm×78.1µm×86.7µm at three different stages of a leaching experiment: 1) saturated with water, 2) injection of 60 g/L potassium iodine (KI) solution after 6 min, and 3) injection of the solution after 78 min. Soil macropore structure and tracer distribution were reconstructed at the five positions. Pore volume, pore hydraulic radius, fractal dimensions, and lacunarity functions in both 2-D and 3- D were computed. The results showed distinct macropore and flow patterns at the five positions in the soil column. The continuous earthworm burrows and root channels were very active in solute transport because of its high continuity and low tortuosity. Positive logarithmic trends were found between fractal dimensions in both 2-D and 3-D and the fraction of interests. The lacunaritiy functions varied along the soil depth and among different features (e.g., macropore network and tracer distribution). Both macropore network and tracer distribution showed higher degree of cluster at the subsurface, which were associated with the biopores and preferential flow and transport. The nearly linear lacunaritiy function of tracer distribution, especially at 78 min, indicated the existence of a self-similarity. The 3-D and 2-D lacunarities were comparable for the data used in this study. Compared with fractal dimensions, lacunarity appears to be more sensitive to the structure difference and more effective in differentiating macropore and flow patterns.