NGWA Conference on Groundwater in Fractured Rock and Sediments: Alphabetical Content Listing

Advances in the Understanding of Fracture Occurrence and the Importance of Fractures in Groundwater Flow Systems

An Integrated Approach to Identify Predominate Flow Zones within Fractured Shale Bedrock

Kristen Musgrove
An integrated approach was used to identify predominate flow zones within a fractured shale bedrock to understand its hydrogeologic relationship to the underlying regional sandstone aquifer system. The site is located in the Midwest and the fracture shale unit is impacted by a release of chlorinated solvents. The integrated approach included bedrock coring using roto-sonic techniques, borehole geophysics, and interval packer testing for groundwater quality and hydraulic head measurements.

Bedrock core borings were advanced using roto-sonic drilling to assess the potential presence of chlorinated solvent contamination within the fractured shale bedrock. An extensive borehole geophysical suite, including flowmeter logs, was run in each bedrock core boring, and is capable of identifying fracture flow zones. Fracture flow zones have the ability to provide principal groundwater flow pathways that constitute the discrete aquifer units of a bedrock formation (Michalski and Britton 1997). Borehole geophysics offers a suite of tools able to characterize bedrock flow zones. Additionally, interval packer testing was conducted during the advancement of each core boring. The interval packer testing provided an assessment of groundwater quality within discrete bedrock zones. The interval packer testing also provided an opportunity to measure hydraulic heads within each interval sampled. When completed across the site, the core borings provided sufficient hydrogeologic characterization to connect potential groundwater flow zones.

Multiple lines of evidence support the findings that contaminants of concern (COCs) were not present below the overburden water-bearing zone. These multiple lines of evidence include: no detection of COCs in the wells screened in the shallow bedrock water-bearing zone; geophysical data and hydraulic head measurements indicating limited potential for vertical (downward) migration of COCs in groundwater; and negligible to very low hydraulic connection among depths within the core borings.

Deciphering Flow Conditions and Evaluating Sustainability of Groundwater in Fractured Crystalline Bedrock

Meredith Metcalf
The crystalline bedrock is a critical water resource for the state of Connecticut, as well as many other mountainous areas. Characterizing the hydrogeology of the fractured rock is difficult given the complex nature of fracture systems. Thus, groundwater basins are assumed to be the same as drainage basins derived from surface topography and the estimated rate of recharge to the bedrock is typically unknown. This research provides a new approach for estimating groundwater sustainability in fractured rock that entails synthesis of pre-existing well data into a comprehensive database that permits defining bedrock groundwater drainage basins and flow for use in estimating recharge and usage. The method was tested in the Coventry Quadrangle, Connecticut and entailed the use of more than 2500 wells integrated into a geographic information system. Results indicate that groundwater drainage basins of the fractured rock do not correspond to drainage basins derived from surface topography, and groundwater recharge estimates for delineated groundwater fractured rock basins are lower than those reported in other studies. Additionally, groundwater recharge and usage for each drainage basin were estimated and the sustainability of each basin was determined by taking the difference between these estimates. Temporal analysis of well parameters indicated a decrease in well yield by approximately 20% and the depth to water declined.

Fracture Occurrence and Connectivity in a Siliclastic Aquifer Near Public Supply Wells in Southern Wisconsin

Kenneth R. Bradbury
Historically, most hydrogeologists and water-supply engineers have treated siliclastic “sandstone” aquifers as classic porous media for the purposes of well design and groundwater protection. Recent discovery of human viruses in deep public supply wells in southern Wisconsin suggests that rapid flow pathways occur in the sandstone aquifer system there. Geophysical, flowmeter, and optical logs of multiple boreholes, combined with hydraulic testing, show that the confined aquifer system contains a network of interconnected fractures. Major near-horizontal fractures occur at discrete stratigraphic positions and conduct flows exceeding 100 gallons per minute in the vicinity of pumping wells. In places, these fractures account for most of the transmissivity of the aquifer and could facilitate rapid movement of contaminants. Steeply dipping fractures that could act as vertical pathways also occur but are less frequently observed. Two significant near-horizontal fracture zones have been incorporated into a new regional groundwater flow model, and model results demonstrate their influence on groundwater flow paths and velocities.

Understanding the presence and significance of these fracture networks is important for appropriate analysis of well vulnerability and for wellhead protection planning. Fracture networks help explain the apparent vulnerability of deep wells to viruses and other contaminants and can significantly influence the calculated zones of contribution used for wellhead protection plans.

Improved Estimates of Hydraulic Conductivity and Specific Storage from Straddle Packer Tests in Fractured Sandstone

Pat Quinn
A method is presented for obtaining depth-discrete values of specific storage (Ss) for short test intervals in single boreholes in fractured rock. The test equipment, described by Quinn et al. (2012), is used to conduct four types of hydraulic tests in each test interval: constant head step tests, slug tests, and pumping/recovery tests. Hydraulic conductivity (K) values are obtained from the constant head step test and slug tests, avoiding errors due to non-Darcian flow. The pumping tests, which have a duration of one half to two hours, provide values for hydraulic diffusivity (K/Ss) obtained by the standard Theis-based curve fitting method. To minimize the usual uncertainty in the hydraulic diffusivity values inherent in the curve fitting methods, the Darcian K values obtained from the constant head step test and slug tests are used to constrain K in the curve fitting procedure to obtain the best value for specific storage (Ss). Care is taken to account for wellbore storage, most influential during the first few minutes of pumping tests and to avoid boundary effects that may influence late-time results. This procedure was applied in 1.5 m intervals in boreholes between 200 and 800 ft deep in a sandstone aquifer which has high fracture density and rock matrix porosity of 5-15%. In this study K values of the sandstone ranged from 3.8 to 7.5 × 10-5 m/s and the Ss values were on the order of 10-6-10-5 m-1, consistent with values reported in the literature. Improved estimates of K and Ss provide much more insight into the hydraulic properties of the fracture network, and allow improved calculation of hydraulic apertures and effective fracture porosities for estimating average linear groundwater velocities. This study illustrates the benefits of conducting four different types of hydraulic tests to gain confidence in both K and Ss.

Novel Geochemical Tools for Characterizing Hydraulically Active Fractures

Amy Hudson, REM
Strong preferential flow paths, very low porosity, low storage, and high degrees of flow anisotropy make rock and aquifer properties of fractured bedrock very different from classically studied aquifer materials (uncemented and cemented clastic sediments). As a result, classic aquifer characterization and analysis tools more appropriately applied to porous media do not apply to fractured rock aquifers. The characteristics of hydraulically active crystalline rocks are primarily controlled by weathering and structure. Several studies have identified the importance of residence time, rather than physical structure as a controlling mechanism of weathering reactions, and that the relative analyte concentrations released vary depending on the system. This suggests that a correlation can be made between weathering product concentrations and the presence of advective flow within a fractured bedrock aquifer.

This project used modeling to simulate weathering under varying residence times and flow lengths to provide expected reaction product concentrations for correlation with the laboratory testing results. Specifically, the goal was to determine the minimum contact time of the fracture water with the mineral surface to result in measureable concentrations of weathering products. For this study biotite was the mineral used to represent the weathering of the fracture surface, and resulting weathering products of potassium, ferric iron oxide, sodium, magnesium, and silicic acid. Geochemical modeling of the weathering reactions and transport of the weathering products was completed in one dimension using CrunchFlow (Steefel 2009), a multicomponent reactive flow and transport model. The CrunchFlow simulations focused on the signatures of the water being transported by the fracture and the mineral weathering on the fracture surface. The results of this modeling effort were able to confirm the importance of residence time. However, further modeling is needed using kinetically controlled reactions in combination with the advective transport to better understand the timing of constituent release.

Some Solutions to Necessary, but Risky, Open Hole Well Development

Carl Keller
Boreholes in fractured rock are often drilled for two reasons, water supply or contamination assessment and remediation. Both uses benefit from development of the well. Well development of a cased hole with a screen and filter pack is easy. Development of an open uncased hole in fractured rock is much more problematic. In an open hole, the usual development methods can collapse the hole, trap tools in the hole, or lead to other extremely costly consequences. However if the transmissivity distribution is an important parameter to the situation, and it usually is, the transmissivity of the formation cannot be measured with the fractures plugged with drill cuttings. Casing the hole is not an option if access to the entire formation is needed. This paper describes some of the relative merits of several development techniques in open holes. Some measurements are provided showing the effect of insufficient well development in open holes. The practical limitations of some common methods are calculated. The production of large amounts of contaminated water can be costly. A new technique using a flexible liner is added to the list of development methods. Transmissivity changes during the development process should be measured to help determine when the development procedure is sufficient to the purpose of the borehole. Several such measurement methods are described. Examples are provided of the results of development in sandstone and limestone formations using flexible liners.

The Geometry and Hydraulic Properties of Fractured Rocks from Particle Transport Measurements in Pumped Wells

Malcolm Anderson
Recently developed hydraulic equations that govern the transport of particles through fractured and coarse granular aquifers provide great insights into the geometry and hydraulic properties of fractured and porous rocks. These equations have many applications, but are particularly useful in the analysis of fractured rocks using elapsed time-particle measurements during single well pumping tests. The turbidity responses determine whether the fracture network is planar or orthogonal, whether it is uniform, the (flow path) capture radius of the well, rock matrix filtration, and the hydraulic conductivity, transmissivity and percentage of flow through the porous matrix of the rock. Particle size distribution measurements against time can be used to determine the flow velocity in the fractures at any radius from the well from the mobile particle sizes. These measurements can be combined with classical time-drawdown data that reveal the single fracture, karst conduit, dual permeability components and homogeneity of the fracture network to compile a single, double or triple permeability hydraulic conceptual model with detailed characterization of the fracture network and solution widened features. As particles are mobile in rock fractures but are filtered out by porous media and porous matrix blocks; and mobile particles transport many harmful compounds that are otherwise immobile through adsorption, attachment, and ion exchange; bacteria, heavy metals, carbon compounds with low partition coefficients (e.g., PAHs) and radionuclides; resolution of the fracture network and the particle transport therein is the key to success. It is therefore essential to undertake particle measurements as a key data set, especially as the cost is low and the benefit is high, to determine the characteristics of the fractured media; the rates and constraints of the particle transport in the fractured media, and the transport constraints and rates of the compounds transported by the particles.

.

Utilizing Pumping-Induced Reverse Water Level Fluctuations to Evaluate Fracture Connectivity in a Siliciclastic Aquifer System

Christopher Gellasch
Reverse water-level fluctuations (RWFs), a phenomenon in which water levels rise briefly in response to pumping, were detected in monitoring wells in a fractured siliciclastic aquifer system near a deep public supply well. The magnitude and timing of RWFs provide important information that can help interpret aquifer hydraulics near pumping wells. A RWF in a well is normally attributed to poroelastic coupling between the solid and fluid components in an aquifer system. In addition to revealing classical pumping-induced poroelastic RWFs, data from pressure transducers located at varying depths and distances from the public supply well suggest that the RWFs propagate rapidly through fractures to influence wells hundreds of meters from the pumping well. The rate and cycling frequency of pumping is an important factor in the magnitude of RWFs. The pattern of RWF propagation can be used to better define fracture connectivity in an aquifer system. Rapid, cyclic head changes due to RWFs may also serve as a mechanism for contaminant transport.

Advances in the Understanding of Fracture Occurrence and the Importance of Fractures in Groundwater Flow Systems (cont,)

Contaminant Transport in a Fault Zone Aquifer

Robert M. Bond, P.G.
Clusters of domestic and irrigation wells along the trace of the Farmhaven Brook Fault Zone attest to the highly permeable nature of the fault zone aquifer. The convergence of commingled groundwater plumes into the systematic network of steeply-dipping en echelon extensional fractures and bedding-parallel fractures aligned subparallel to the fault zone created the need to better understand the structures and contaminant transport in this fractured aquifer. The bedrock in this New Jersey study area is the Passaic Formation of the Newark Basin consisting of interbedded shales, siltstones, mudstones, and sandstones. We used several approaches to better characterize fracture networks, including data from outcrops; borehole televiewer, caliper, natural gamma and flow metering logging, bedrock topography, tracers, aquifer hydraulic tests, seismic refraction velocity cross-sections and 3D-visualization.

Two north-south trending, steeply-dipping en echelon faults, lying approximately 1000 feet apart, are the focus of this presentation. Vertical gradients over the 4000-foot VOC plume are generally downward with the exception of these steeply-dipping en echelon breaks where strong upward gradients and even flowing artesian conditions prevail. VOC-impacted bedrock groundwater is transported up into the shallow overburden aquifer over one of these breaks, creating a contaminated island leading to vapor intrusion investigations, one residential mitigation and surface water issues. The upwelling of “clean” bedrock groundwater on other fault segments and temporal variations in vertical head in bedrock control the position of the commingled overburden plumes. Two seismic refraction lines were run across one of the en echelon breaks and a portion of the fault ramp and then modeled to better understand the fracture networks and buried fluvial deposits. The analysis of spatial and temporal geochemistry trends were also used to create the conceptual fracture flow model. The installation of additional overburden and bedrock wells, aquifer testing, and downhole geophysics in early 2013 will supplement our research.

Cross-Hole Heat Tracer Tests to Assess Groundwater Flow in Fractured Rock: Successes and Lessons Learned

Peeter Pehme, Ph.D.
Characterization of groundwater flow through fractured rock is essential for assessing and predicting contaminant plume migration but is both difficult to perform and costly, while often yielding uncertain interpretations. Solute tracer tests are informative, but subject to regulatory approval. Furthermore, introduction and extraction of water as well as the use of conventional wells or boreholes with long open sections disturb the very environment being characterized. The combination of high sensitivity temperature techniques and removable borehole liners to eliminate cross-connection provide an improved means of using temperature as an innocuous tracer. Heat can be introduced under natural hydraulic gradients to assess ambient groundwater through fractured rock and provide a guide to a potentially complex flow/fracture pattern and the process can be easily repeated. This describes tests wherein heat is used as a tracer in fractured rock, the successes achieved, and lessons learned.

Three boreholes (each 150m deep, in a triangular pattern of ~10m sides, lined with FLUTeä sleeves) in Cambridge, Ontario, were used for the tests. The “natural” gradient in the area is controlled by several distant municipal pumping wells. The entire length of the upgradient borehole of the trio was heated for 5d in one test and 14d in a repeat experiment. Temperatures along the length of all three boreholes were measured regularly over the test periods.

In the first test an initial breakthrough arrived in one well after 49hrs of testing, resulting in a conservative estimate of groundwater velocity as 3.4m/d. The initial 0.050C¢ª response increased in magnitude and expanded to additional fractures/depths over the following 41hr, highlighting a complex distribution of numerous flow conduits. The repeat of the test did not provide as dramatic a result, presumably due to changes in the ambient groundwater flow direction; however, valuable insight was gained for planning additional cross-hole heat tracer tests.

Fracture Identification Using Low-Cost CR-39 Detectors

Bill Brandon
It has long been recognized that preferential flow, primarily through interconnected fracture networks, is critical to understanding the exchange of contaminants between bedrock and the shallow subsurface in water and gas phases. However, the difficulty in identifying discrete fracture-controlled transport pathways adversely impacts all phases of site characterization, modeling, and remediation at fractured bedrock sites. Boreholes do not always penetrate critical fractures, and fracture connections between boreholes are difficult to determine. There is, therefore, a critical need to develop inexpensive yet robust tools to better understand the location, geometry, and extent of significant fractures. Utilizing 222Rn as an indicator of fracture-controlled transport pathways may represent an inexpensive assessment tool for identifying fractures in some bedrock settings. Our research evaluated the applicability of using low-cost readily-available CR-39 222Rn detectors to locate and map conductive pathways (i.e., active fractures and fracture networks) in the subsurface at a shallowly buried granitic bedrock upland located at Fort Devens, Massachusetts. Unique to this approach is the development of a new rapid deployment port that allows the CR-39 detector to be placed within or above bedrock, while restricting atmospheric 222Rn from reaching the detector. For this study, a power auger was used to penetrate thin overburden deposits in order to place the detector-casings just above the top-of-bedrock surface, generally less than 5 feet below grade. Our research demonstrated that by deploying a series of these inexpensive detector-casing combinations statistically significant measurements of the 222Rn flux could be collected. These measurements allowed for accurate delineation of the most communicative fractures (i.e., actively transporting water and gasses) at the study site. The CR-39 data was then used to validate the existing site conceptual model. In contrast to other available fracture-locating techniques, deploying our technique, in some cases, potentially offers improved X-Y spatial resolution at significant cost savings.

Prediction of the Macroscopic Properties of Fractured Porous Media

Pierre Adler
A major concern in the study of fractured media is the prediction of their properties such as the percolation of the fracture network and the macroscopic permeability from the few field quantities which are easily measurable.

Our approach is based on the systematic use of the excluded volume of fractures. When the percolation threshold of the fracture network and the macroscopic permeability are plotted as functions of r', defined as the number of fractures per excluded volume, they become independent of the fracture shapes, which is a decisive simplification for the applications.

r' can be estimated from measurements performed on intersections of fracture networks with lines, planes, and galleries. These intersections are visible on outcrops, cliffs, quarries, wells, and tunnels. Some remarkable relations hold whatever the fracture shapes if they are convex.

Applications of this approach to real cases will be discussed.

During the presentation, the meshing of the fractured medium and the discretization of the equations will be detailed. This methodology can be applied to arbitrary fracture network geometries, and to arbitrary distributions of permeabilities in the porous matrix and in the fractures.

Characterizing and Remediating Large Dilute Plumes in Fractured Media

John N. Dougherty, PG

Application of a Conceptual Site Model for Groundwater Contamination in a Bedrock Municipal Well System

Lisa Campbell
This paper describes how a conceptual site model (CSM) was developed and used by EPA to identify the sources of volatile organic compounds affecting municipal water supply wells completed in the bedrock aquifer at a Superfund site in Puerto Rico. The site is underlain by Cretaceous age volcanic rocks of the Pre-Robles Formation, which consist mostly of massive volcanic breccia. Chemical weathering has resulted in an overlying saprolite layer, which is a major water-bearing unit that stores water and provides recharge to the underlying bedrock aquifer. The saprolite is semi-confined by the overlying silty clay soils in the overburden.

The CSM was developed over the course of a multi-staged investigation, which included the following activities:

  • Soil investigations of potential source areas, including discrete-depth soil sampling through the overburden.
  • A hydrogeologic investigation using borehole geophysics, heat pulse flow meter logging, and packer testing to determine the frequency and orientation of bedding, fractures, and joints and to identify water-bearing zones in six bedrock boreholes. These data were used to design a multilevel well for each location.
  • Investigation of the saprolite zone, which had been found during the bedrock drilling activity to be a major water-bearing unit. The investigation included installation and sampling of 11 single-screened wells to determine the extent of contamination in this zone.
  • An evaluation of the groundwater-surface water interaction in the river downgradient of the contaminated supply wells. Stream gauges were installed and measured, and surface water and sediment samples were collected.
  • Preparation of cross-sections using the results of this program to develop the bedrock CSM.

These data were compiled and evaluated to develop and refine the CSM of the groundwater flow system. The CSM was then used in conjunction with the soil data to identify two sources of the groundwater contamination.

Delineation of Trichloroethene and Related Contaminants in Weathered and Unweathered Sedimentary Rock, NAWC, New Jersey

Daniel J. Goode
The U.S. Geological Survey, Toxic Substances Hydrology and National Research Programs, in cooperation with the U.S. Navy and the Strategic Environmental Research and Development Program, is conducting multidisciplinary research on the distribution and fate of chlorinated volatile organic compounds (CVOCs) in weathered and unweathered fractured sedimentary rocks that underlie the former Naval Air Warfare Center (NAWC), West Trenton, New Jersey. Trichloroethene (TCE), a dense non-aqueous-phase liquid, was used in industrial processes at NAWC from the mid-1950s to the mid-1990s, and high concentrations have been detected in groundwater samples. Below the weathered zone, groundwater flow and transport of CVOCs take place locally along bedding-plane fractures and thin fissile or laminated strata in gently-dipping mudstones of the Triassic-age Lockatong Formation of the Newark Basin. Concentrations of CVOCs in most rock-core samples from the shallow weathered zone exceed limits of detection. In contrast, concentrations in core samples from deeper unweathered strata are generally below detection. However, high CVOC concentrations are present in isolated unweathered-rock samples from highly fractured dark carbon-rich mudstones and fractured dark-gray laminated mudstones. Consistent results from five coreholes, along with hydraulic and water-quality characterization, suggest a conceptual model for the migration of CVOCs from or near the land surface into the aquifer. Prior to pump-and-treat remediation (P&T), groundwater flow was generally horizontal in the high-permeability weathered zone, and the bulk of contaminant mass was shallow. CVOCs diffused into the weathered and fractured rock for decades. After P&T began in 1995, primarily using bedrock wells open below the weathered zone, CVOCs were pulled downward from the weathered zone into sub-cropping, dipping high-permeability bedding plane fractures. The data indicate diffusion of TCE and other CVOCs from deeper fractures only penetrated a few centimeters into the unweathered rock matrix, possibly due to sorption of CVOCs on the organic material in the dark mudstones.

Developing a Remedy for a Large Complex Chlorinated Solvents Plume in a Fractured Rock Environment

John N. Dougherty, PG
This presentation will detail site characteristics important for development of a remedy at the Cayuga County Groundwater Contamination Site, which extends from Auburn, New York, 7 miles to the southwest to the village of Union Springs near Cayuga Lake. Understanding of these characteristics—groundwater flow, contaminant distribution within the system, and natural attenuation processes—informed development of the proposed remedy, including division of the site into four areas: the former industrial site and three downgradient areas. The remedy includes the ongoing remediation of the source, currently being conducted by the owner and NYSDEC, and remedial actions proposed by EPA for the three downgradient areas. For Area 1, closest to the source, enhanced anaerobic biodegradation is proposed. Monitored natural attenuation is proposed for Area 2, which extends southwest from Area 1, and for Area 3, which extends to Union Springs.

The site is underlain by a shale and carbonate fractured bedrock aquifer system. Past waste disposal practices at the source have caused high levels of volatile organic compound (VOC) contamination in groundwater underlying the facility and immediately downgradient. VOC concentrations decrease significantly downgradient of the former facility. To characterize the bedrock aquifer system, EPA worked closely with USGS. A series of multilevel wells were installed near the source area and downgradient towards Union Springs. Borehole geophysical tools were used to characterize the hydrogeology in these boreholes and to support design of the multilevel wells. The owner of the source area has also conducted extensive investigations, including monitoring well installation and characterization of biotic and abiotic degradation in the aquifer. EPA, the USGS, and the owner have shared data and coordinated groundwater level monitoring and sampling events.

Influence of Fracture Fabric and Gradients on Contaminant Migration at the Savage Well Superfund Site

Andrew Fuller, PG
Historical use of tetrachloroethene at the Savage Well Superfund Site in Milford, New Hampshire impacted both the overburden and bedrock aquifers underlying the site, including the nearby municipal overburden water supply well. Although the municipal well was taken offline, residences to the north of the source area are serviced by private bedrock water supply wells. Site geology consists of transmissive glacial outwash and a discontinuous till unit overlying fractured crystalline bedrock.

A multi-phased approach including well installations, borehole geophysics, packer interval sampling, and a pumping test was employed to develop a conceptual site model for the fractured bedrock aquifer at the site. Borehole geophysics and a 76-hour pumping test were utilized to evaluate fracture frequency, strike, and dip, identifying hydraulically active fractures within each monitoring well, and determining the anisotropy of the bedrock aquifer.

The investigation identified a competent bedrock fabric with a limited number of hydraulically active fractures that were moderately to steeply dipping (median 66º), with a predominant strike to the north-northeast. The pumping test data displayed a similar trend with a clear north-northeast anisotropy to the cone of depression. However, static hydraulic head distribution indicated easterly flow across the aquifer, almost perpendicular to the predominant strike of the fractures and anisotropy. To further complicate interpretation, hydraulic conditions at the site had been modified over the years as a result of the installation of a slurry wall barrier and groundwater extraction system in the overburden. It was not initially clear whether the fracture fabric or the hydraulic gradient had greater control over contaminant migration. Determining the actual direction of contaminant migration was essential to determining whether residential wells to the north of the site were at risk. This presentation will discuss the investigation, the final conceptual model for the site, and the fate of the residential wells.

Investigation of Alternative Groundwater Sampling Systems for Use in Fractured Rock Aquifers

Philip T. Harte
Groundwater sampling in open boreholes tapping fractured bedrock aquifers is particularly challenging because of mixing of water within the borehole from multiple fractures. To alleviate this problem, a packer-pump assembly can be employed to isolate flow from discrete fractures or fracture sets. Groundwater is then extracted at a rate commensurate with flow produced by the identified fracture. While this procedure has long been recognized as a standard procedure to collect samples from discrete fractured zones, it can be time consuming and logistically difficult.

This study is investigating alternative (without the use of a packer assembly) sampling systems that can be used to extract representative groundwater samples from boreholes adjacent to discrete fractures with less equipment and at a reduced cost. A patent-pending system has been developed that utilizes hydraulic containment to extract representative groundwater samples in open boreholes. Representative samples are obtained by minimizing mixing of borehole water with fracture water. Therefore, a key component of the system is the confirmation of favorable borehole flow patterns during sampling. A variety of tools have been used to track borehole flow patterns, including flowmeter and novel tracer deployment procedures. Preliminary results indicate that fracture water can be effectively isolated from stagnant borehole water with the new sampling system.

Real-Time Assessment of VOCs in Fractured Bedrock Using Innovative Core Discrete Fracture Network (DFN) Techniques

Seth E Pitkin
Following an initial program of site investigation using a variety of High Resolution Site Characterization (HRSC) techniques in accordance with the U.S. EPA Triad framework, chlorinated solvent impact was identified beneath an active manufacturing site within unconsolidated alluvial deposits at the base of the alluvial deposits and upper margins of underlying shale bedrock.

In order to determine vertical contaminant distribution and extent, a second stage of works was implemented comprising detailed assessment of the shale bedrock via the Core DFN approach. This included collection of cores for geological logging, field screening (with a photoionization detector) and on-site pore water extraction using Microwave Assisted Extraction (MAE) to rapidly obtain circa 450 crushed rock samples for on-site VOC analysis using a GC/MS.

The use of the MAE equipment enabled the investigation to be completed using a dynamic Triad style approach and provided near real-time on-site analyses of bedrock matrix contamination data that was used to progressively refine the investigation scope in a sustainable manner.

The results of the fractured bedrock assessment provided a refined and detailed Conceptual Site Model showing that the contaminant plume migrated laterally through the weathered shale profile and the interface with deeper, fresher fractured shale bedrock provided a “barrier” to significant vertical contaminant migration.

The results of the chemical analysis identified two trichloroethene source zones and indicated that significantly greater contaminant concentrations were present within pore water samples (up to 1620 mg/L), than the dissolved phase concentrations detected in samples collected from monitoring wells had initially shown (up to 40 mg/L total VOCs), reflecting a typical mass distribution for fractured rock with the greatest concentrations present within the rock matrix.

The refined conceptual site model was instrumental in the development of a technically appropriate and cost effective sustainability-led remediation approach, which was subsequently implemented at the site.

Stringfellow Superfund Site – Characterization, Remediation, and Modeling of Groundwater Impacts

James M. Finegan, PhD, PG, CHg
The Stringfellow Superfund Site in Riverside, California, overlies an aquifer system comprised of alluvium, weathered bedrock, and unweathered fractured bedrock.  Impacts to groundwater extend approximately 5 miles downgradient from a box canyon containing the source area, through an alluvial paleovalley, and ultimately toward a river.  In the source area where over 34 million gallons of industrial waste were deposited, groundwater impacts are present primarily within bedrock.  This condition persists within the canyon, with transport transitioning primarily to the alluvial paleovalley downgradient of the canyon.  However, even within bedrock, transport appears to be associated with fracturing controlled by antecedent structural components that initiated the formation of the canyon and paleovalley.

Remediation to control migration of VOCs and perchlorate includes pump-and-treat systems in each of the aquifer units; recent enhancements to these systems exhibit likely improvements in plume control.  Following detection of perchlorate in 2001, remedial investigations for the Stringfellow Site have focused on control of this compound, although additional perchlorate sources (e.g., Chilean fertilizer and quarry blasting chemicals) have also been identified. Isotopes are being used to attempt source differentiation.  Bench and pilot scale in-situ bioremediation testing for perchlorate reduction has also been performed.  Overall effectiveness was not well supported in bedrock, in part due to competing electron acceptor pathways and potential difficulty in distribution of electron donor.  A bromide tracer test performed with this study was used to evaluate transport velocities.

A three-dimensional numerical model of groundwater flow and perchlorate transport in this system downgradient of the source area was developed in 2003 and has subsequently been updated and refined to assist in predictions of solute transport and evaluation of remedial alternatives.  In addition, a three-dimensional site model has been developed to depict the aquifer system and help elucidate the complex hydrogeologic interactions and transport between these units.

Uranium Occurrence and Arsenic Variability in Private Well Water in Southeast New Hampshire

Marcel Belaval
Uranium, derived from geologic sources, is a frequently occurring contaminant in groundwater in New England’s bedrock aquifers. While recent studies have helped to quantify the presence of arsenic in New England’s fractured rock aquifers, less is known about the distribution of uranium in these aquifers. The goal of this ongoing study is to characterize the occurrence and analyze the spatial distribution of uranium in private well water through randomized sampling of private wells in southeastern New Hampshire. In addition, the study examines temporal variability in arsenic, since the majority of wells used in this study were also sampled 10 years ago as part of an arsenic occurrence and distribution study in the same area. To date, approximately 225 samples have been analyzed for arsenic, iron, manganese, lead, and uranium. Samples were collected and submitted by private well owners in the study area. Initial uranium occurrence results show that about 47% of wells had detectable (> 1 ug/L) concentrations of uranium, but less than 2% of the samples exceeded the U.S. Environmental Protection Agency’s (EPA) Maximum Contaminant Level (MCL) of 30 ug/L for uranium. Comparisons of field-filtered samples to unfiltered samples, which were collected for a subset of wells, showed little variability in concentrations for arsenic, lead, and uranium, indicating that these analytes were fully dissolved in the groundwater. Comparisons between arsenic measurements from 2002 and 2012 show similar concentrations: 19% of samples were above the MCL of 10 ug/L in 2002, compared to 17% in 2012. However, individual wells showed slightly more variability: only 80% of samples with arsenic greater than 10 ug/L in 2002 were still above 10 ug/L in 2012.

Diffusion and Reaction Processes in Rock Matrices

Investigating Well Connectivity and Reactive Surface Area in a Sandstone Bedrock Using Ionic Tracers

Matthew W. Becker, Ph.D.
Pump and treat remediation of bedrock groundwater systems is often impacted by an uneven sweep of the formation between injection and pumping wells.  Poor sweep efficiency may lead short circuiting of flow and treatment or, in the opposite extreme, poor water and reagent circulation in the impacted fractures.  We describe a field approach to measuring sweep efficiency using multiple ionic tracers.  Non-reactive anionic tracers are used to trace the water while reactive cationic tracers are used to probe the fracture surface.  Assuming the cation exchange capacity of the fracture surface is relatively constant, separation of cation breakthrough curves is a function primarily of available surface area.  We present results from a series of ionic tracer experiments conducted in a single bedding plane fracture in a low-matrix-porosity sandstone formation.  Tracer breakthrough was collected among five wells offset by 7 to 14 m.  Breakthrough was interpreted using numerical transport modeling that accounts for cation exchange.  We compare the tracer estimates swept surface area with estimates obtained from ground penetrating radar images of saline solution circulated between the same well pairs.  We find that, even though tests were performed in a single fracture, swept fracture surface area is highly variable among well pairs.

Investigation of Hexavalent Chromium Matrix Diffusion at a Sedimentary Bedrock Superfund Site

Steven W. Chapman, M.Sc., P.Eng.
It is now well known that diffusion driven mass transfer between fractures and the rock matrix has major implications on plume fate and transport at contaminated sites overlying fractured bedrock aquifers.  Restoration of these sites poses a major technical challenge since most available technologies only address the contaminated water in fractures and back diffusion from the matrix to the fractures causes rebound which acts to slow cleanup and reduce effectiveness of the remedial actions.  A study of matrix diffusion and immobilization reactions of hexavalent chromium [Cr(VI)] is being conducted as part of the remedial investigation of contamination emanating from a Superfund site (a former electroplating facility) in New Jersey situated within the Newark Basin.  Previous studies within this fractured sedimentary bedrock basin and elsewhere have primarily focused on dense non-aqueous phase liquid (DNAPL) sites with VOC contamination.  This is the first study focused on chromium, which is a contaminant of concern at many Superfund sites nationwide.  This study involved collecting a detailed profile of Cr(VI) concentrations in the bedrock matrix from continuous cores at one location within the plume.  Besides diffusion into and out of the matrix, chemical reactions may also act to immobilize the Cr(VI) via conversion to immobile Cr(III) in the matrix, which is assessed in this study.  Supporting investigations included borehole geophysics, transmissivity profiling, and installation of a multilevel well for depth discrete hydraulic head and groundwater sampling for comparison with rock core data.  This presentation will provide an overview of the study approach, field and laboratory investigation and analytical techniques, interim results and implications.  Ultimately this study is expected to provide a framework for understanding fracture network characteristics and transport pathways and matrix diffusion and reaction processes controlling Cr(VI) transport and fate, allowing a more accurate evaluation of risk to receptors, remediation technologies and cleanup timeframes.

Matrix Diffusion Effects on Nitrate Fate and Transport in Prince Edward Island's Sedimentary Bedrock Aquifer

Amanda Malenica, BSc.
The province of Prince Edward Island is unique in Canada because 100% of the population is reliant on groundwater. Island wide studies have shown increasing nitrate levels associated with intensive agriculture, a major concern both from a human health standpoint and surface water impacts (eutrophication, fish kills). A multicomponent research study is being conducted at four sites focused on nitrate behavior in the fractured bedrock aquifer. As part of this study, effects of matrix diffusion are being evaluated including impacts on nitrate persistence with reduced loading as a result of changes in management practices. The Discrete Fractured Network (DFN) approach is being applied to this problem, which recognizes that much contaminant mass in fractured sedimentary aquifers may reside in the pore water of the rock matrix due to diffusion. The matrix (which has high porosity, typically 5-20% in sandstone) has very large storage capacity compared to the fractures (which have very low porosity, typically 10-3 to 10-5). A key study component is collecting detailed profiles of nitrate distribution from subsampling of continuous cores from boreholes ranging in depth from 80 to 200 feet with matrix pore water analyses for various analytes. Other components include borehole geophysical methods, transmissivity profiling, and design and installation of multi-level groundwater monitoring systems for detailed understanding of hydraulic head and groundwater concentrations. Discrete zones of interest will be monitored long term for temporal changes in nitrate and other analytes. This research will aid in the understanding of the role of matrix-stored nitrate and attenuation mechanisms, and ideally in linking groundwater nitrate fluxes to agricultural management. Ulimately this will improve conceptual models and help agricultural researchers better understand controls on the fate of nitrate under different loading scenarios leading to more scientifically defensible decisions on agricultural best management practices and water supply protection.

Measurement of the Spatial Distribution of Heat Exchange Using Fiber Optic-Distributed Temperature Sensing

Adam Hawkins
Highly channelized flow in fractured geologic systems has been credited with early thermal breakthrough and poor performance of geothermal circulation systems.  An experiment is presented here in which the effect of channelized flow on fluid/rock heat diffusion is measured.  Hot water was circulated between two wells in a single bedding plane fracture and the elevation of rock matrix temperature was measured using fiber optic Distributed Temperature Sensing (DTS).  Between wells with good hydraulic connection, heat transfer followed a classic dipole sweep pattern.  Between wells with poor hydraulic connection, heat transfer was skewed toward apparent regions of higher transmissivity (or larger aperture).  These results are consistent with hydraulic and tracer tests, as well as ground penetrating radar imaging, that shows a heterogeneous distribution of transmissivity. The results suggest that flow channeling can have a significant impact on heat transfer efficiency even in single planar fractures.

Quantifying Three-Dimensional Matrix Diffusion Effects on Plume Front Retardation

Barry Brouwers
It is well accepted that matrix diffusion effects on transport of contaminants through fractured media results in an effective retardation of plume migration.  Discrete Fractured Network (DFN) modelling representations for plume migration in fractured rock are typically accomplished using a two-dimensional approach.  While this approach is the limit of what is practical at the plume scale for many sites that have ubiquitous fracturing, it is expected that the effective plume retardation within such simulations will underestimate the rate of retardation that occurs in the field.  To better quantify the degree of enhanced retardation within the natural environment, three-dimensional discrete fracture modelling was undertaken.  For this approach, the discrete fracture flow and contaminant transport is accomplished by superimposing two-dimensional plane elements onto the three-dimensional mesh that represents the rock matrix, and incorporating a detailed level of distretization away from this fracture-matrix block interface to represent steep and variable diffusion profiles.  Insights gained through this detailed numerical modelling provide a better understanding toward the prediction of plume-scale migration rates that appropriately account for matrix diffusion and the apparent plume front retardation.

Keynote Address: Flow and Transport Properties of Fractured Bedrock Aquifers in the Vertical Direction

Kent S. Novakowski, Ph.D.

Measuring Mass Flux/Mass Discharge and Groundwater Flow Velocity in Fractured Media

A Diffusive/Advective Model of a Pore Water Transition Front in a Borehole Connecting two Fractures

Edwin A. Romanowicz
We report the findings of a diffusive/advective model of specific conductivity profiles in an uncased borehole.  The well in this study is 140 m deep, fully penetrating the Cambrian Potsdam Sandstone at our field site (Flat Rock, West Chazy, New York).  Over a period of several months we have documented the movement of a specific conductivity transition front.  This transition front is characterized by a change in specific conductivity from 75 to 260 µS·cm-1.  The transition front moves between two fractures intersecting the well at depths of 24 and 34 m.  We have not observed this occurrence in neighboring wells. Even though some of these wells have fractures common with the well in this study.

We speculate that this transition front is driven by changes in the contribution of groundwater flow to and from the borehole between the two fractures.  Relative percent concentrations of major cations (Ca, Mg, Na, and K) of borehole-water samples collected above and below the transition front suggest different water sources for the fractures.   The water from the deeper fracture shows influence of increased Mg, consistent with dolomitic units in the Potsdam.  Preliminary analysis of the transition front suggest that average vertical flow between the fractures was 0.45 m·sec-1 (6 mL·min-1). 

The extended time over which this transition front has been monitored offers us the opportunity to model temporal changes in the flow to and from the borehole through these fractures.  We are developing a diffusive/advective model to characterize the movement of this transition front over time in response to changes in water flux through the fractures.  We will model several different configurations to test multiple hypotheses explaining the transition front.  The model is being developed using Stella ®.

The Fractured Rock Passive Flux Meter: A New Tool for Characterizing Flux in Fractured Systems

Kirk Hatfield
Fractured rock settings pose substantial economic and technical challenges both to the characterization and remediation of DNAPL source zones.  The objective of this project is to demonstrate and validate the fractured rock passive flux meter (FRPFM) as new technology for measuring the magnitudes and directions of cumulative water and contaminant fluxes in fractured rock aquifers.  The sensor consists of an inflatable core that compresses a reactive fabric against the wall of a borehole and to any water-filled fractures intersected by a borehole. The reactive fabric is designed to intercept and retain target groundwater contaminants (e.g. PCE, TCE, DCE, and VC); in addition, the fabric releases non-toxic tracers, some of which visibly indicate active fracture location, aperture, orientation, and direction of fracture flow along a borehole, while others quantify cumulative groundwater discharge within the fractures.

Field demonstration tests are ongoing at the Naval Air Warfare Center (NAWC) in West Trenton, NJ. Along with demonstrating the capabilities of the FRPFM, the tests are also being used to compare multiple technologies including, optical televiewer, acoustic televiewer, high resolution temperature logging, and borehole dilution tests. The technologies are being evaluated to generate a collaborative standard operating procedure to optimally identify flowing fractures, determine flow direction, and quantify both water and contaminant mass flux within fractured systems.

Panel: Bedrock Remediation Current Perspectives

Michael B. Smith

Poster Session

Measuring Contaminant Mass Flux and Groundwater Velocity in Fractured Rock Aquifer Using Passive Flux Meters

Diana M. Cutt, PG
The development of cost effective techniques for the assessment and remediation of contaminated fractured rock aquifers has been identified as a science priority in EPA Region 2. This project, funded by EPA’s Regional Applied Research Effort (RARE) Program, will research the ability of an innovative tool – a fractured rock passive flux meter (FRPFM) - to measure groundwater flow velocity and mass flux in a fractured bedrock setting and to compare the results to current technology. The study will be implemented at a Superfund site in Puerto Rico where bedrock aquifers are an important source of drinking water. The goal is to research the ability of the FRPFM and standard passive flux meter (PFM) to measure groundwater flow velocity and mass flux in fractures in a fractured bedrock setting and to compare the results to current technology. A PFM consists of a cylinder of activated carbon impregnated with tracer compounds that can be installed in a well screened in unconsolidated deposits to measure groundwater flow velocity and contaminant mass flux. The FRPFM is an experimental system developed by the University of Florida. Using this tool, the activated carbon fabric material of the meter is placed against the bedrock in the borehole wall and uses packers to isolate different intervals of the open borehole. The research borehole will be geophysically logged and packer samples will be collected before installation of the flux meters. Both the FRPRM and the PFM are proposed to be installed in the research borehole at three or four depths so that the data generated can be compared to information obtained by geophysical logging and packer sampling from this and other boreholes at the site. Use of both PFM designs in the research borehole will allow comparison of the results generated by the different methods.

Potpourri

High-Resolution Investigation of Vapor Intrusion in Fractured Sedimentary Rock

Daniel Carr, PE, PG
The presentation will detail methodologies and findings of investigations and testing for vapor intrusion associated with chlorinated ethenes in siltstone and sandstone located in the mid-Atlantic U.S.  There is little scientific or regulatory precedence for assessment of vapor intrusion in fractured soil or rock environments, as most of the empirical experience is derived from work at granular soil sites, with little relevance to fractured rock settings.  Fractured rock may be distinguished from a granular soil setting through lower gas-filled porosity to support transport and the possible influence of fracture orientation.

 Our presentation will focus on how the historical site conceptual model, developed from investigations conducted with long open bedrock boreholes, was refined through higher resolution investigation and pilot testing of remedy enhancements.  Application of discrete fracture network (DFN) style investigation techniques (Parker 2007) included high-resolution logging of core and open boreholes, physical and chemical testing of core samples, and multi-level monitoring of water and subsurface gas under stressed and unstressed conditions. We will present estimates of fracture porosity and effective hydraulic aperture derived from this testing compared against effective porosity estimates derived from gas and aqueous phase tracer testing, and apparent specific yield estimates derived from hydraulic testing.

 The data derived from this testing support that upward vapor migration in the subsurface is limited by the presence of sparsely-fractured aquitard intervals and a predominance of near–horizontal, bedding-plane parallel fracturing.  The patterns of fracturing as indicated by logging of core and boreholes area are supported by the findings of multi-level monitoring of water and gas under unstressed conditions and under applied stresses from relatively large–scale, dual-phase extraction testing.  Given the predominance of near-horizontal fracturing, we found that it is possible to establish a vacuum field underlying many acres of land through vacuum dewatering of fracture networks.

Quarry Influences in Limestone Environments

David Ketcheson
Paleozoic limestones in Southern Ontario are laterally extensive and dip slightly into the Michigan Basin.  The competent limestone formations near ground surface tend to be characterized by laterally extensive conductive fracture planes created in part by glacial unloading.  These fracture planes provide local ground water resources for residents; but are not so well developed to yield significant water resources for large municipal supply.

The influence measured at various quarry developments cut into these limestone plains in Southern Ontario was assessed.  EPM estimates poorly reflect the inherent complexity and heterogeneity of a fractured rock environment.  This study evaluates the similarities and differences that exist at various quarry operations across a portion of Southern Ontario based on more recent instrumentation of various working sites.  The evaluation of actual site data is considered to provide a much more discerning evaluation than is possible through theoretical estimates.  The fractured bedrock setting was examined with the intent of identifying critical characteristics which tend to define the consistency in the outward influence on ground water flow as measured at numerous quarry sites in Southern Ontario.  The study also provides practical insights into the development of effective operational monitoring programs at these facilities which will safeguard domestic supplies in proximity to such operations.

Using Deep Bedrock Well Logs to Constrain Stratigraphic and Structural Problems in Vermont

Jonathan Kim
Recently, the Vermont Geological Survey (VGS) received a multi-year grant from the U.S. Dept. of Energy to study the deep geothermal potential of Vermont. One task required by this grant is to acquire groundwater temperature logs for 20 deep (³500’) bedrock wells, which are distributed among the tectonic belts of Vermont. The VGS partnered with Geology faculty at SUNY at Plattsburgh to conduct this logging, using their Mt. Sopris equipment. In addition to the measurement of temperature profiles and the calculation of geothermal gradients, conductivity, gamma, and caliper logs are routinely run.

             At present, we have logged 4 deep wells which are completed in: 1) Red sandstone and interbedded dolostone of the Middle Cambrian Monkton Formation, part of the hanging wall of the Ordovician Champlain Thrust; 2) Rift-clastic metasedimentary rocks of the Neoproterozoic Pinnacle Formation; 3) Schist and gneiss of the Mesoproterozoic Mt. Holly Complex; and 4) Cambrian-Ordovician carbonates and shales of the footwall of the Hinesburg Thrust, after drilling through phyllites of the Neoproterozoic Fairfield Pond Fm. of the hanging wall. 

            Because logs traverse a continuous section of bedrock that is not accessible from surface exposures, they are particularly valuable for examining stratigraphy and structure at depth. Hand-measured surface sections of the Monkton Formation have revealed meter-scale cyclical shallowing-upward packages that reflect Cambrian sea level fluctuations. We seek to correlate the sedimentary package cyclicity with that seen in the gamma logs.

The Hinesburg Thrust is poorly exposed at the ground surface and domestic well logs are used to determine its geometry at depth. The gamma log for well #4 has pronounced inflections which (top to bottom) are interpreted as the base of the water-bearing zone in hanging wall phyllites, the thrust zone where phyllites change abruptly to carbonates, and the footwall transition from carbonates to shales.

Remediation in Fractured Rock Environments – Effectiveness and Innovation

Beth L. Parker, Ph.D.

Advances in Remediation of Fractured Bedrock Using In Situ Thermal Treatment Technologies

Kevin Leahy, PhD, C.Geol
Following an extensive programme of site characterisation using a variety of High Resolution Site Characterisation techniques, chlorinated solvent impact was identified beneath one area of the site discussed.  The majority of the contaminant mass was within the matrix of saturated bedrock; the most challenging environment to successfully perform in-situ remediation given the confined and fractured nature of the subsurface.

A risk reduction strategy of mass recovery to the extent technically, practically feasible and sustainable was agreed with the regulatory authorities and following an options appraisal, soil and groundwater remediation was undertaken using thermally enhanced Soil Vapour Extraction (SVE) to liberate contaminants from the bedrock.

Whilst several full scale steam injection projects have been undertaken in the UK, since the first application in 2005, SVE has traditionally been undertaken within either a naturally occurring unsaturated zone or one created via dewatering.  At this site the aquifer was found to be highly transmissive and confined by low permeability clay, making both of these options technically unsuitable.  Operation of the remediation system was therefore focussed upon 1) heating the clay from beneath to artificially increase its permeability and 2) development of a ‘steam bubble’ to allow vapour recovery through a zone created by boiling the groundwater.

The results demonstrate considerable success of the above approach with total mass removal calculated at circa 1,100kg after 12 weeks of system operation.

The application of this remediation approach in a complex fractured bedrock setting provides confidence that a similar approach could be taken on other sites where remediation has previously been viewed as technically unachievable.

Characterization and Modeling Approach for Matrix Diffusion and Remedial Alternatives Evaluation in Fractured Sedimentary Rock

Steven W. Chapman, M.Sc., P.Eng.
Detailed site investigations and numerical modeling was conducted to evaluate transport and fate of chlorinated solvent contamination in a fractured sedimentary bedrock aquifer at a Superfund site in New Jersey. Investigations followed the Discrete Fracture Network (DFN) Approach including transmissivity profiling, borehole geophysics, hydrogeophysical logging and intra-borehole flow testing, continuous coring with detailed rock matrix sampling, installation of multilevel wells providing head profiles and depth-discrete groundwater data, and integrated pumping tests to evaluate contaminant mass discharge.  This information provided a framework for understanding fracture network characteristics, transport pathways and matrix diffusion and reaction processes that strongly influence contaminant distribution and fate, which is critical to assessing rates of plume front migration, risks to downgradient receptors, including  municipal well fields, and assessment of remedial alternatives.  The field data provided the basis for numerical modeling which was conducted in two stages: 1) an equivalent porous media (EPM) flow model (MODFLOW-2005) was used to evaluate the bulk groundwater flow system and contaminant transport paths under different historic aquifer stress conditions; then 2) results of the EPM model informed a discrete fracture network (DFN) model (FRACTRAN) for rigorous assessment of flow and transport including effects of matrix diffusion and other processes such as sorption and degradation that control plume transport and fate in fractured sedimentary rock.  The combined model results were used to further develop the conceptual site model and evaluate remedial strategies, and showed contaminant transport occurs in a well-interconnected network of fractures and that matrix diffusion and other processes cause strong plume attenuation such that future risks to municipal well fields should be negligible. Modeling also showed futility of further source remediation efforts, which would have minimal impact on the plume over practical timeframes.  Results of the study were used in support of a Technical Impracticability (TI) waiver for the site.

TCE Remediation Using Electrical Resistance Heating in a 90-Foot-Thick Rock Sequence

Mark Kluger
Remediation practitioners have known for years that electrical resistance heating works exceptionally well in tight soil matrices. Sedimentary rock treatment has become rather common in the last five years; however, until now, no one has applied ERH in a substantially thick layer of rock. A consultant in the northeastern United States, working closely with TRS Group, a company that provides ERH services, decided that a 90-foot thick sedimentary rock sequence at a site in eastern Pennsylvania would be an ideal setting for the technology and recommended ERH in the Remedial Action Plan submitted to the Pennsylvania Department of Environmental Protection. PADEP agreed and approved ERH as the remedy. This marks the first time that anyone has applied the technology in such a thick rock unit.

The presentation will discuss ERH theory, the application in rock, results and lessons learned.

Remediation in Fractured Rock Environments – Effectiveness and Innovation (cont.)

Beth L. Parker, Ph.D.

Changes in Groundwater Biogeochemistry Caused by Bioaugmentation Remediation in a Fractured Sedimentary Rock Aquifer

Thomas E. Imbrigiotta
An in situ bioaugmentation experiment to stimulate the degradation of chlorinated ethene compounds was initiated at the Naval Air Warfare Center in October 2008 by injecting emulsified soybean oil as an electron donor and a consortium of bacteria.  Groundwater flow from the injection well is predominantly toward a pump-and-treat withdrawal well located about 45 meters (m) downgradient.  Baseline geochemical conditions were established during the six months prior to injection by sampling 5-8 times for chlorinated ethene compounds, bacteria, inorganic geochemical constituents, and dissolved gases from the injection and withdrawal wells and from two multi-level observation wells located along the flow path.  Following injection, the same constituents were monitored in the same wells 14-18 times over the next four years to assess the effects of the bioaugmentation.

From 2008-2012, substantial reductions (two orders of magnitude) were found in trichloroethene (TCE) concentrations in groundwater from the injection well and in one of the sampling intervals from the monitoring well 18 m downgradient, with lesser  reductions (one order of magnitude) in one of the sampling intervals of the monitoring well 30 m downgradient.  These sampling intervals are coincident with mudstone beds that are the primary conduits for flow between the injection and withdrawal wells.  In water from these same sampling intervals, increased concentrations (1-2 orders of magnitude) of dechlorinating bacteria, cis-1,2-dichloroethene, vinyl chloride, ethene, and chloride were found, indicating that the decrease in TCE was primarily caused by reductive dechlorination.

Concentrations of lactic acid, one of the electron donors added, decreased to pre-injection levels in the injection well within two and a half years.  Concentrations of acetic acid, a breakdown product of the soy bean oil, remained two orders of magnitude above pre-injection levels, in the same well, indicating that degradation of this electron donor may still be ongoing four years post-injection. 

Pilot Test Evaluation of High-Pressure Jet Injection for In situ Remediation of Low-Permeability Zones

Neal Durant, Ph.D.
In situ remediation of chlorinated solvents in clay till and saprolite is challenging because low-permeability constrains reagent delivery, and solvent diffusion into the matrix substantially limits the rate of treatment. Pneumatic and hydraulic fracturing can enhance delivery of treatment agents; however, their radius of influence (ROI) is limited and primarily controlled by the natural fracture networks. Here we present the results of two pilot studies using jet injection, a novel 10,000 psi hydraulic fracturing method to create large hydraulic fractures in in clay till and saprolite. Pilot test results showed greater  ROI than conventional fracturing methods, the potential to bypass natural fractures, and feasibility of implementation on a rapid and flexible drilling platform.

A pilot test was completed using jet injection to deliver zero-valent iron from two wells into clay till at a test site in Taastrup, Denmark. The resulting fracture network was mapped by excavating the 18m x 24m x 8m test plot. The injection achieved an ROI of 6 to 7m, farther than previous hydraulic fracturing in similar clay tills. However, the loss of kinetic energy caused by jetting through the well casing and surrounding grout resulted in short circuiting into the natural fracture system rather than creation of new fractures. To improve performance, a second pilot test combining direct push technology (DPT) and jet injection was implemented in South Carolina saprolite. Customized tooling was fabricated for DPT rods that distributed all of the kinetic energy of the water jet into the formation. The DPT injection of water and dye at 10,000 psi resulted in the formation of fracture cavities up to 2.1 m long and 90 cm wide. The innovative combination of DPT and jet injection provides a flexible, rapid, and promising method to achieve a large injection ROI that bypasses natural fractures in low-permeability zones.

Remediation of Fractured Bedrock Containing Light Non-Aqueous Phase Liquid (LNAPL)

Sean R. Carter, PE
Light non-aqueous phase liquid (LNAPL) and elevated volatile organic compounds (VOCs) were detected in groundwater at an active gasoline station in Niagara Falls, New York.  The station was closed and the USTs, fuel lines, and one set of fuel dispensers were removed.    Surficial geology consists of 5 to 9 feet of fill material or native clay overlying highly fractured, gray to light gray limestone.  Groundwater is present in bedrock at approximately 11 feet bgs and flows in a fracture-oriented southeasterly direction.  Monitoring wells contained LNAPL up to 2.5 feet in thickness and groundwater BTEX concentrations as high as 82 milligrams per liter (mg/L).  A Remedial Action Plan was developed to remove LNAPL and reduce groundwater VOCs using multiple technologies in a phased approach.

Submersible top-loading pneumatic pumps were installed in multiple wells.   LNAPL and groundwater was pumped to an on-site treatment system and discharged to a sanitary sewer.  The system operated for 47 months, treated nearly 500,000 gallons of groundwater, and reduced site-wide LNAPL thickness to less than 0.10-feet.

The remediation system was then augmented to include vacuum-enhanced groundwater pumping to increase well yield, improve hydraulic control, and extract subsurface vapor-phase VOCs.  Financial restrictions associated with permitting necessitated the premature deactivation of the pump and treat portion of the system after 18 months of operation.  The system extracted 473 pounds of VOCs, and was successful in removing any remaining LNAPL and reducing groundwater BTEX concentrations to below 0.5 mg/L. 

To complete remediation, the injection of oxygen gas into 2 injection wells screened in bedrock fractures was tested in mid-2012, with eventual full scale implementation into 15 injection wells.  Groundwater dissolved oxygen has increased to the target level of 5 mg/L and the system will be operated for an estimated 18 months.

Role of Geologic Faults on Contaminant Plume Morphology and Implications for Remediation

David S. Lipson, Ph.D., PG
Groundwater remediation in fractured rock can be challenging, impracticable, or even impossible due to structural controls including faults, folds and joints. Such features complicate evaluation, selection, and implementation of effective remediation strategies. Failure to account for these controls can result in remedies that are ineffective and can exacerbate problems. We designed groundwater remediation programs at three contaminated sites where complex plume morphologies were controlled by geologic faults. The remediation programs took into account the structural geology present at these sites and are in various stages of implementation and monitoring.

To properly conceptualize the sites, we used groundwater modeling, bedrock coring, downhole televiewers, and high-resolution groundwater monitoring, as well as fracture analysis, hydraulic testing, geochemical evaluations, and analysis of rock samples. This information was used to conceptualize the sites, determine the dominant controls on complex plume morphology, evaluate failure modes, and design remedial programs.

Results showed that geologic faults can exert significant control on the fate, transport, and remediability of groundwater contaminants, and resulted in complex plume shapes and migration patterns that would not be predicted assuming homogeneous and isotropic conditions. Not considering geologic structures during site characterization, remedial design, and implementation would have resulted in the implementation of ineffective remedial strategies.

Surface Water/Groundwater Interaction in Fractured Rock Environments

Coliform Bacteria in Bedrock Groundwater: Insights from the NJ Private Well Testing Act Database

Thomas Atherholt, Ph.D.
It has now been over 10 years since enactment of the NJ Private Well Testing Act (Sept 2002).  As of Sept 2012 the NJDEP has accumulated water quality data from over 80,000 wells, and over 93000 samples (including wells sampled more than once).  Each sample was analyzed for a variety of parameters, including coliform bacteria.  The database has some limitations with respect to data quality, but there is no sampling bias with respect to water quality and very little bias with respect to sampling time or location.  Analysis of the coliform bacteria data revealed a number of factors that affected the bedrock detection rates.  These factors include; the laboratory and analytical method used, the number of times a well is analyzed, season, precipitation, and geology.  Based on a single sample per well, bedrock detection rates (percent positive wells) were 20% for total coliform bacteria and 3% for fecal coliform or E. coli bacteria.  Wells located in sedimentary rock were more vulnerable to coliform contamination than wells located in igneous or metamorphic rock and wells with low-pH water (3-6) were more vulnerable than wells with higher-pH water (7-10) regardless of rock type.  Wells located in areas with thick glacial sediment were less vulnerable than wells located in areas with little or no overburden.  Both temperature and precipitation appeared to affect monthly coliform detection rates, which are higher during warm months.

DNAPL or Dissolution? Mercury Transport to Riverbed Fractures at a Former Chlor-Alkali Facility

Jennifer Lambert, P.G.
Significant quantities of mercury migrate from the former Brown Company’s Chemical Plant (a chlor-alkali facility) into the adjacent Androscoggin River. The former Chemical Plant is located in Berlin, New Hampshire immediately downstream of the hydroelectric Sawmill Dam. The local bedrock consists of quartz monzonite gneiss and granitic pegmatite with lenses of chlorite schist, all of which are extensively folded and irregularly fractured.

Between 1999 and 2004, approximately 170 pounds of mercury and mercury-contaminated debris were removed from the riverbed. Mercury, both as liquid droplets and semi-solid and solid amalgam, continues to accumulate in bedrock fractures along the river almost 50 years after the cessation of the Chemical Plant operations. A CERCLA Remedial Investigation was completed to help identify the mercury source location and possible mechanisms for mercury transport and discharge to the River. 

Local hydrology is complicated by several factors: an impermeable source area cap; a partially confining slurry wall in the overburden upgradient of the primary source area which diverts overburden groundwater into the bedrock; and the schedule of water releases from the dam, which may not coincide with local precipitation. The means of mercury transport to the surface may be the simple migration of non-aqueous phase mercury through bedrock fractures driven by hydraulic gradients.  However, geochemical conditions in the subsurface may allow the dissolved transport of mercury with subsequent precipitation in liquid and amalgam forms once exposed to surface water conditions.  The Plant’s past operational practices have resulted in unusual geochemistry in groundwater underlying the former Chemical Plant’s footprint, including elevated pH and high concentrations of organic compounds and metals.  Defining these conditions and their effect on mercury speciation are critical in understanding the conditions for mercury transport and guiding future remedial actions.

Equivalent Representative Fractured Network for Modeling Groundwater Flow

Shaul Sorek
Fractured media feature, e.g., dead end pathways, lack of flow interconnections and preferential infiltration paths. These are, somewhat, not in line with the assumptions allowing the implementation of a continuum based balance partial differential equations. We thus developed a lumped parameter modeling (LPM) approach to assess the groundwater flow patterns through the fractured granites of Porto Alegre, Southern Brazil. Observations consist of stationary groundwater levels mostly localized at part of the domain, and geometrical properties of surface lineaments over the 30 by 35 Km study domain. Evaluation of their hydraulic head values revealed the existence of isolated lineament clusters, which could not be detected a priori and generated not acceptable flow or no-flow zones. We thus subdivided the domain into a map of Representative Elementary Area (REA) cells, chosen so that the ratio of lineament lengths summation over that of the cell area remains practically unchanged between two consecutive subdivisions. The actual surface lineaments were replaced by an Equivalent Representative Fracture Network (ERFN). The evaluated steady state hydraulic head was then compared between two LPM solutions referring to: 1) the intersection points of the ERFN; 2) nodes each being the locus of all ERFN intersection points within the same former REA cell. Computation addressed by the latter solution was significantly less intense. Both approaches lead to hydraulic head isolines consistently similar to those interpolated on the basis of the observed groundwater levels. One available evident of a Nitrate polluted well and its plausible contamination source within the study domain, show that it is oriented with the predicted flow direction along the ERFN.

 

Estimating Discharge of Chlorinated Volatile Organic Compounds from Contaminated Fractured Rock to a Stream

Pierre J. Lacombe
Concentrations of trichloroethylene (TCE), cis-1,2-dichloroethylene (cDCE), and vinyl chloride (VC) and stream base flow discharge measurements were collected at 48 stations in Gold Run, a small brook that passes over contaminated fractured bedrock in West Trenton, N.J.  Surface water-quality data were collected at each station up to 7 times during 1984-98, prior to closure of the Naval Air Warfare Center (NAWC), a former Navy jet engine testing facility adjacent to Gold Run. Water samples were collected at each station up to 48 times during 1999-2012, after closure of the NAWC. Base flow discharge measurements were collected at 12 stations during 2007-12. Mean concentrations of TCE, cDCE, and VC in the brook prior to closure ranged from less than detection to 520, 7,000 and 2,300 micrograms per liter (mg/L) respectively. After base closure, accompanied by contaminated soil excavation, installation of a pump and treat system, and monitored natural attenuation, mean concentrations of TCE, cDCE, and VC in the brook ranged from less than detection to 320, 560, and 37 mg/L, respectively. New Jersey Department of Environmental Protection Agency regulates TCE and VC in surface water with maximum permissible concentrations of 1 and 0.08 mg/L, respectively. Base flow measurements range from zero to 284 liters per minute (L/min). Concentrations of TCE, cDCE, and VC decrease downstream primarily as a result of volatilization, decreasing by 50 percent over every stream reach of 211, 211, and 106 meters respectively. Based on mean concentrations and base flow, 19.8 kilograms per year (kg/yr) of original TCE discharged to Gold Run from an area of highly contaminated bedrock at the NAWC since closure compared to 280 to 6,200 kg/yr prior to closure. This result suggests that contaminated soil excavation, contaminate groundwater withdrawals, and natural attenuation has substantially reduced offsite discharge of contaminants via Gold Run.

The State of the Art in Borehole Geophysical Tools and Methods for Site Characterization

John N. Dougherty, PG

A Photometric Logging Probe for Dilution Logging in Fractured Bedrock Aquifers

Frederick L. Paillet
Dilution logging is capable of precisely indicating inflow depth points while measuring extremely low rates of borehole flow, but requires cumbersome procedures to condition the fluid column, and introduction of distilled water or brine may not be acceptable to regulators on proposed environmental studies.  A photometric probe, the HRTFN Fotometer, performs dilution logging by measuring the concentration of an environmentally harmless food color dye.  The measurement section of the probe uses a LED light source tuned to the exact absorption of the dye so that concentrations as low as 10 mg/l can be used in dilution experiments.  The probe is also fitted with temperature and fluid resistivity sensors, and has a back-scatter optical detector for characterizing the concentration of suspended particles in the fluid column.  The photometric probe was tested in a bedrock borehole in New York known to contain weak ambient downflow expected to be slightly above the lower measurement limit (about 0.10 l/min) of the heat pulse flowmeter.   HP flowmeter measurements made less than an hour before the testing of the photometer indicated a downflow rate of 1.21 l/min with measurement scatter of about 0.10 l/min. An initial column dye concentration of about 60 mg/l was then established.  Repeat profiling of the column with the photometer showed the column diluted by inflow at 24 m in depth, eventually sweeping dye from the column down to an outflow depth at 34 m after 3 hours.   The flow model fit to the photometric data indicated a downflow rate of 1.25 l/min with an estimated error of +/- 0.03 l/min.   These results demonstrate that photometric dilution logging can be effectively used as a substitute for brine dilution in the detection and characterizing of inflow zones in bedrock boreholes where brine dilution is not acceptable and distilled water replacement is impractical.

Demonstrating Three-Dimensional Electrical Resistivity Imaging in Fractured Rock to Characterize Fractures and Monitor Amendment Injections

Judy Robinson
Rutgers University, the U.S. Geological Survey, and Pacific Northwest National Laboratory are demonstrating the use of 3D cross-borehole electrical resistivity imaging (ERI) for improving the characterization of fractured rock aquifers and the monitoring of amendment treatments that biologically stimulate reductive dechlorination of chlorinated ethenes.  Seven, 4-inch diameter, boreholes have been drilled at the Naval Air Warfare Center (NAWC) in West Trenton, New Jersey, which is a mudstone fractured rock site contaminated with trichloroethylene, cis-1,2-dichloroethene, and vinyl chloride.  Interpretations of borehole geophysical logs including caliper, optical televiewer, acoustical televiewer, gamma, and ambient/transient heat pulse flowmeter determined local fracture locations, their hydraulic significance, and connectivity between boreholes.  Cross-hole hydraulic testing identified bedding-plane parting fracture zones and possible fault zones that are the dominant hydraulic pathways.  In a synthetic simulation based on the NAWC site, we found that constraining ERI inversions with this type of fracture zone information from geophysical logs and hydraulic testing substantially improved the resolution of fractures and the prediction of amendment transport away from the boreholes.  A field demonstration of ERI has been designed where 3D electrical resistance measurements were acquired from electrodes in all seven boreholes.  To test whether limiting the preferential current pathway introduced in a fluid-filled borehole will have a significant effect in resolving fracture zones in numerical modeling, highly resistive inflatable packers have been installed between two or more electrodes.  ERI measurements collected during a doublet tracer test will be used to: demonstrate the high-resolution capability of electrical resistivity in fractured rock, suggested in the synthetic studies, and to design an amendment injection for remediation.  Our goal is to demonstrate that ERI using structural constraints in numerical modeling improves characterization of fractures and spatial distribution of amendment treatments.

Determining Flow Conditions in Crystalline Bedrock Wells Using Dissolved Oxygen as a Tracer

Dariusz Chlebica
A novel approach to locate transmissive fractures and ascertain borehole flow conditions in fractured bedrock wells is presented that uses dissolved oxygen (DO) as a tracer. The low background DO level in a wellbore is elevated by circulating water through showerheads or injection of compressed air. Subsequently, the DO levels are monitored with time to ascertain ambient flow conditions or a slug test is performed to simulate pumping conditions. Downhole DO profiling is used to locate inflowing fractures, stagnant zones and ascertain wellbore flow rates. The method was tested in two 94.5 m deep wells in fractured crystalline bedrock (Hebron Gneiss) on the campus of the University of Connecticut, Storrs, CT.  The transmissive fractures identified corresponded to fractures observed in televiewer logs and were comparable to flow meter results. The relative benign nature of dissolved oxygen along with the low-cost and limit logistics of deployment make this method advantageous in drinking water wells or where such wells are nearby.

Fractured Bedrock Aquifer Characterization Using Borehole Geophysical Logging and FLUTe Multilevel Well Systems

John N. Dougherty, PG
The presentation will address challenges associated with designing and installing monitoring wells screened in the proper water-bearing intervals within a fractured bedrock aquifer at a Superfund site in Puerto Rico. Contamination has been detected at the site in municipal supply wells that draw from the bedrock aquifer. A site conceptual flow model was developed using published geologic data, topography, and surface water drainage features. The model was used to identify drilling locations along the likely groundwater flow path(s) between potential source areas and the supply wells. After completion of each borehole, mechanical caliper, natural gamma, acoustic televiewer, optical televiewer, fluid temperature and fluid conductivity logs were run. These data were compiled in WellCAD software and were used to identify potential water-bearing zones for heat pulse flow meter logging under ambient and pumped conditions. Logging results were used to identify groundwater sampling points in the water column above and below water entry and exit points.

The EPA Region 2 low-flow sampling method was used to collect groundwater samples from target depths in each borehole. The samples were analyzed for volatile organic compounds (VOCs). To determine transmissivity with depth, drop tests using blank borehole liners from Flexible Liner Underground Technology (FLUTe) were completed in six of the seven boreholes. Logging data, VOC groundwater sample results, and transmissivity data were evaluated collectively in WellCAD software to identify multilevel monitoring intervals for each borehole. Seven multilevel well designs, including precise monitoring intervals, and the WellCAD file were provided to the multilevel well manufacturer to ensure proper construction of the Water FLUTe® systems. The results of groundwater sampling and water level monitoring from the multilevel bedrock wells will be used to update the conceptual flow model. The model supports ongoing site characterization and efforts to confirm the source of contamination affecting the municipal wells.

High Resolution Hydraulic Profiling and Groundwater Sampling Using FLUTe System in a Fractured Limestone Setting

Gry Sander Janniche, Ph.D.
Characterization of the contaminant source zone architecture and the hydraulics is essential to develop accurate site specific conceptual models, delineate and quantify contaminant mass, perform risk assessment, and select and design remediation alternatives. This characterization is particularly challenging in deposit types as fractured limestone. The activities of a bulk distribution facility for perchloroethene (PCE) and trichloroethene (TCE) at the Naverland site near Copenhagen, Denmark, has resulted in PCE and TCE DNAPL impacts to a fractured clay till and an underlying fractured limestone aquifer/bedrock. A wide range of innovative and current site investigative tools for direct and indirect documentation and/or evaluation of dense non-aqueous phase liquid (DNAPL) presence were combined in a multiple lines of evidence approach. One scope of the investigations was to evaluate innovative investigation methods for characterization of the source zone hydrogeology and contamination, including FLUTe system hydraulic profiling and Water-FLUTe multilevel groundwater sampling,  in fractured bryozoan limestone bedrock. 

High resolution hydraulic profiling was conducted in three cored boreholes, placed within a 970 ft2 (~90 m2) area, and Water-FLUTes were installed with 12-13 sampling screens in each borehole. Hydraulic profilling by FLUTe liner system provided information with highere discretization than other traditionel methods, and supported the individual design of Water-FLUTes for multilevel groundwater monitoring, sampling (under two flow conditions) and analysis. Coring for discrete subsampling was a challenge in the limestone, due to core-loss and potential DNAPL loss caused by high drilling water pressure. Hence, the water-FLUTe data proved to be an essential link in the source zone characterization. The results from the high resolution hydraulic profiling and from the Water-FLUTe multilevel sampling will be presented as well as the experiences obtained.

Understanding Emerging Contaminant Transport and Fate in Groundwater Systems

Cumulative Frequency Analysis of Fractured Bedrock in Conceptual Site Models and Remedial Design

Kevin Leahy, PhD, C.Geol
Cumulative Frequency Analysis (CFA) of geological structures is a technique commonly deployed on borehole core and in outcrop by the hydrocarbon industry to elucidate geofluid flow through fractured bedrock. Typically, CFA forms the primary input data for complex quantitative fracture network models used by exploration, reservoir engineering and production geologists. The CFA technique is not commonly deployed in the environmental sector, where structural data analysis is mainly qualitative or, more rarely, in discrete feature network models.

The CFA technique was applied as part of a thermal remediation project in the UK, where trichloroethene impacts were identified at depths up to 18 m in fractured Carboniferous mudstones and sandstones. Careful logging of bedrock structures and the degree of bedrock weathering was undertaken. Graphical field sketch logs were utilised to capture structural data, with care taken to avoid the inclusion of mechanically-induced fractures.

A CFA methodology was developed using industry-standard tools to process structural data and provide critical information on contaminant fate and transport. This showed that, at this particular site, there was no coincidence of logged fractures with the distribution of contaminant plumes. Rather, the contaminant plume was seen to be migrating within a finely anastomosing mesh observed in the highly weathered mudstones, with no penetration into structures within the fresh bedrock. This may be explained by the frequent observation of a clay fill within the structural features at the interface zone of fresh and weathered bedrock. The revised conceptual site model was used to optimise the thermal remedial design to focus on the identified contaminant source zones and migration pathways, greatly reducing the depth and lateral extent of rock to be treated. This resulted in a lower cost, more rapid and more sustainable remedial outcome that ultimately recovered 1.6t trichloroethene. The applicability of CFA to environmental projects is discussed.

Naturally-Derived Arsenic in Fractured Slate: Influence of Bedrock Geochemistry, Groundwater Flow, Reduction-Oxidation and Ion Exchange

Jonathan Kim
Groundwater hydrochemical and bedrock geochemical analysis indicates that elevated As (up to 155 ppb) in the Taconic slate aquifer system of southwestern Vermont is controlled by four main factors: (1) the presence of black slates rich in arsenian pyrite (with 200 – 2000 ppm As); (2) release of As via the oxidation of As-rich pyrite; (3) reducing conditions — the highest As values occur at Eh < 200 (and pH > 7); and (4) physical hydrogeological factors that foster low Eh and high pH, particularly long groundwater flow paths and low well yields (i.e. high residence time).  Where all four of these factors affect groundwater, 72 % of wells in a zone of distal groundwater flow/low-relief topography exceed the US EPA MCL of 10 ppb and 60% of wells in this zone exceed 25 ppb As.  Where flow paths are shorter and groundwater has higher Eh (i.e. in regions of higher-relief topography closer to recharge zones), only 3 % of wells contain > 10 ppb As and none contain > 25 ppb. Overall, 28 % (50/176) of wells with wellhead elevations between 60 and 245 masl exceed 10 ppb As; only 3 % (2/60) of wells (wellheads) situated between 245 and 600 masl exceed 10 ppb As.  Over the entire aquifer system, 22 % of bedrock wells (52/236) exceed 10 ppb and the mean As concentration is 12.4 ppb.  Strong positive correlations among Fe, SO4 and As in groundwater confirm that dissolution of pyrite is the dominant As source. Positive correlations among SO4, Na and As indicate that, in reducing (Eh < 200) groundwater, Fe(II) is exchanged for Na on mineral surfaces following pyrite dissolution and As remains in solution; in oxidizing groundwater (recharge zones), Fe(II) is oxidized to Fe(III) and the subsequent formation of Fe-hydroxides removes As from solution.

Welcome

William Alley, Ph.D.