Elsevier

Journal of Contaminant Hydrology

Volume 171, 15 December 2014, Pages 1-11
Journal of Contaminant Hydrology

High-resolution delineation of chlorinated volatile organic compounds in a dipping, fractured mudstone: Depth- and strata-dependent spatial variability from rock-core sampling

https://doi.org/10.1016/j.jconhyd.2014.10.005Get rights and content

Highlights

  • Synthesis of rock core CVOCs with detailed lithology and borehole characterization

  • CVOCs were widespread in shallow weathered, highly fractured sedimentary rock.

  • CVOCs were detected only at strata-bound fractures below the weathered zone.

  • Limited CVOC diffusion into unweathered rock may help remediation via amendments.

Abstract

Synthesis of rock-core sampling and chlorinated volatile organic compound (CVOC) analysis at five coreholes, with hydraulic and water-quality monitoring and a detailed hydrogeologic framework, was used to characterize the fine-scale distribution of CVOCs in dipping, fractured mudstones of the Lockatong Formation of Triassic age, of the Newark Basin in West Trenton, New Jersey. From these results, a refined conceptual model for more than 55 years of migration of CVOCs and depth- and strata-dependent rock-matrix contamination was developed. Industrial use of trichloroethene (TCE) at the former Naval Air Warfare Center (NAWC) from 1953 to 1995 resulted in dense non-aqueous phase liquid (DNAPL) TCE and dissolved TCE and related breakdown products, including other CVOCs, in underlying mudstones. Shallow highly weathered and fractured strata overlie unweathered, gently dipping, fractured strata that become progressively less fractured with depth. The unweathered lithology includes black highly fractured (fissile) carbon-rich strata, gray mildly fractured thinly layered (laminated) strata, and light-gray weakly fractured massive strata. CVOC concentrations in water samples pumped from the shallow weathered and highly fractured strata remain elevated near residual DNAPL TCE, but dilution by uncontaminated recharge, and other natural and engineered attenuation processes, have substantially reduced concentrations along flow paths removed from sources and residual DNAPL. CVOCs also were detected in most rock-core samples in source areas in shallow wells. In many locations, lower aqueous concentrations, compared to rock core concentrations, suggest that CVOCs are presently back-diffusing from the rock matrix. Below the weathered and highly fractured strata, and to depths of at least 50 meters (m), groundwater flow and contaminant transport is primarily in bedding-plane-oriented fractures in thin fissile high-carbon strata, and in fractured, laminated strata of the gently dipping mudstones. Despite more than 18 years of pump and treat (P&T) remediation, and natural attenuation processes, CVOC concentrations in aqueous samples pumped from these deeper strata remain elevated in isolated intervals. DNAPL was detected in one borehole during coring at a depth of 27 m. In contrast to core samples from the weathered zone, concentrations in core samples from deeper unweathered and unfractured strata are typically below detection. However, high CVOC concentrations were found in isolated samples from fissile black carbon-rich strata and fractured gray laminated strata. Aqueous-phase concentrations were correspondingly high in samples pumped from these strata via short-interval wells or packer-isolated zones in long boreholes. A refined conceptual site model considers that prior to P&T remediation groundwater flow was primarily subhorizontal in the higher-permeability near surface strata, and the bulk of contaminant mass was shallow. CVOCs diffused into these fractured and weathered mudstones. DNAPL and high concentrations of CVOCs migrated slowly down in deeper unweathered strata, primarily along isolated dipping bedding-plane fractures. After P&T began in 1995, using wells open to both shallow and deep strata, downward transport of dissolved CVOCs accelerated. Diffusion of TCE and other CVOCs from deeper fractures penetrated only a few centimeters into the unweathered rock matrix, likely due to sorption of CVOCs on rock organic carbon. Remediation in the deep, unweathered strata may benefit from the relatively limited migration of CVOCs into the rock matrix. Synthesis of rock core sampling from closely spaced boreholes with geophysical logging and hydraulic testing improves understanding of the controls on CVOC delineation and informs remediation design and monitoring.

Introduction

Remediation of groundwater at fractured-rock sites contaminated with Dense Non-Aqueous Phase Liquids (DNAPL) and dissolved chlorinated volatile organic compounds (CVOC) remains a major challenge. A recent strategic study by the U.S.'s National Research Council (NRC) (2013) noted “The most problematic [hazardous waste] sites are those with potentially persistent contaminants including chlorinated solvents recalcitrant to biodegradation, and with hydrogeologic conditions characterized by large spatial heterogeneity or the presence of fractures.” In fractured-rock settings, especially sedimentary rocks, the rock matrix is a reservoir for contaminants that can act as a long-term secondary source by DNAPL dissolution, back diffusion, and de-sorption (Dearden et al., 2013, Mutch et al., 1993, Parker et al., 2010, Révész et al.,2014).

Characterization of the CVOC mass distribution in fractured sedimentary rocks has recently advanced using rock core sampling techniques, providing a fine scale vertical profile of CVOC mass diffused into or sorbed onto rock strata (Sterling et al., 2005, Parker et al., 2010). Analysis of rock core for CVOCs provides fine-scale delineation of contaminants in and on the rock matrix, and is not affected by possible vertical contaminant migration in the borehole after coring (cross contamination). Dearden et al. (2013) used rock core sampling to identify the CVOC profile into about 4 m of Triassic-age mudstone underlying a surficial aquifer. Chapman et al. (2013) recently presented rock core sampling results from 3 coreholes in fractured sedimentary rocks of the Newark Basin, with depths of about 70, 75, and 120 m, respectively, and horizontal spacings of more than 200 m apart along a flow path.

Our study extends previous interpretation of the rock core sampling results by synthesis with borehole geophysical methods and multi-level water-quality monitoring within a site conceptual model based on a detailed hydrogeologic framework, using closely spaced coreholes of depths of up to 60 m. This approach allows a site-scale characterization of the control of weathering and strata-bound permeability features on CVOC migration and persistence in sedimentary fractured rocks. Our study also provides additional direct confirmation of some of the fine scale features included in current conceptual models of CVOC migration in fractured sedimentary rocks, such as limited diffusion into deep unweathered rock. These results are used to refine the conceptual model for CVOC fate and transport in fractured sedimentary rock aquifers.

In this investigation, we synthesize closely spaced rock core sampling results with multi-level water-quality monitoring, borehole geophysical logging, and hydraulic testing at the former Naval Air Warfare Center (NAWC), West Trenton, New Jersey (Fig. 1). Lacombe, 2000, Lacombe, 2002, Lacombe, 2011 describes monitoring well configurations at the site and interpretation of water levels and contaminant concentrations measured in monitoring wells installed by the U.S. Navy. Industrial use of TCE at the NAWC resulted in DNAPL TCE, and dissolved TCE, cis-1,2-dichloroethene (DCE), and vinyl chloride (VC) in groundwater, and historic contaminant discharges to Gold Run, a culverted stream along the site's southern boundary (Fig. 1). The distribution of peak CVOC concentration reflects source areas where DNAPL TCE was released, and is still present, and migration of dissolved CVOCs towards Gold Run. Since the mid-1990s a pump and treat (P&T) system has minimized discharge of CVOCs beyond the site boundary and to Gold Run (Lewis-Brown and Rice, 2002).

The evolving conceptual site model for contaminant migration at NAWC is based on a detailed hydrogeologic framework (Lacombe, 2000, Lacombe and Burton, 2010). The study area is underlain by mudstones of the Lockatong Formation of Triassic age, of the Newark Basin, one of a series of early Mesozoic Basins in the eastern U.S. in which mudstone, shale, and sandstone aquifers are used for water supply (Trapp and Horn, 1997). At NAWC the mudstone strata are gently dipping to the NW, and a fault along the southeastern boundary of the site separates the Lockatong from the older sandstone of the Stockton Formation (Fig. 1). The detailed lithology of the local mudstones is divided into distinctly colored strata that occur in cycles of deposition: thin (less than 0.2-m thick) black strata containing up to 7% organic carbon by weight (Marjorie S. Schulz, USGS, written comm., 2007; Lebron et al., 2013) and generally highly fractured (fissile), gray mildly fractured thinly layered (laminated) strata, and light-gray weakly fractured massive strata. Several of the high-carbon fissile black beds are herein designated by their respective depths in corehole 43BR at the site: for example the top of a black fissile stratum designated “BlkFis-233” occurs at a depth of 233 ft (71 m) in 43BR.

Below relatively high-permeability weathered and highly fractured strata, groundwater flow is primarily in bedding-plane fractures and thin fissile high-carbon strata, or in fractured, laminated strata in the dipping mudstones, as identified by borehole (Williams et al., 2007) and aquifer testing (Lacombe, 2000, Tiedeman et al., 2010). Large-scale horizontal anisotropy in transmissivity is a well-known characteristic of Newark Basin aquifers, with preferential water-level responses to pumping in the direction of strike of the dipping beds (Herpers and Barksdale, 1951, Longwill and Wood, 1965, Vecchioli, 1965). This apparent horizontal anisotropy is caused by the dipping hydro-stratigraphy, together with the vertical anisotropy of hydraulic conductivity (HK): HK is highest parallel to the dipping beds and lowest perpendicular to beds (Michalski, 1990, Michalski and Britton, 1997, Goode and Senior, 1998, Goode and Senior, 2000, Senior and Goode, 1999, Lacombe, 2000).

Although the precise timing, magnitude, and location of spills of DNAPL TCE are unknown, it is presumed that spills occurred at multiple locations over several decades (Lacombe, 2000). These presumed localized source areas and highly heterogeneous groundwater flowpaths have resulted in highly variable spatial distributions of CVOCs (Fig. 1b) as characterized by water samples from wells with open intervals of generally between 4 and 8 m, at depths of up to about 125 m (Lewis-Brown and Rice, 2002). Despite more than 18 years of P&T remediation, and presumably longer natural attenuation, CVOC concentrations remain elevated at many locations (Fig. 2; Lacombe, 2011; also see Supplementary information figs. SI2 and SI3). TCE concentrations are usually much higher than that of degradation byproducts DCE or VC in wells open to deeper strata (24BR, 36BR, 56BR), whereas degradation byproduct DCE is often the predominant CVOC in wells open to shallower strata (BRP1, 7BR, 15BR). Chapelle et al. (2012) estimated much higher rates of TCE degradation in shallower strata than in deeper strata at the site. DeFlaun et al. (2006) and Révész et al. (2014) observed that the relation of TCE > DCE in highly contaminated deep wells was reversed to DCE > TCE after bioaugmentation.

Section snippets

Methods

New core holes were installed using standard rock core diamond-bit drilling methods. Shallow alluvium was removed by auger, and in a limited number of shallow intervals, tri-cone rotary drilling (not coring) was used for short intervals of highly fractured, weathered rock that could not be recovered in the core barrel. Tap water was used to circulate cuttings, and small amounts of synthetic mud was added to the recirculating fluids when needed to prevent binding of the bit. The coreholes were

Results and synthesis

The site-scale synthesis and stratigraphic correlation of rock core CVOC concentrations are illustrated using closely spaced coreholes in two areas of the site. The results and synthesis are presented in detail for the “West Area” (Fig. 1c). CVOC concentrations in water samples pumped from monitoring wells (or sampled by diffusion bags) as part of the U.S. Navy's site monitoring program and from other USGS research studies are briefly summarized to compare the spatial variability shown in those

Refined conceptual model of CVOC transport

Consistent rock-core sampling results from five coreholes, in synthesis with hydraulic and water-quality monitoring and hydrogeologic characterization, suggest refinements to the conceptual site model for the migration of CVOCs and depth-dependent rock matrix concentrations. Active remediation included P&T from wells starting in 1995 (Lacombe, 2000) and excavation in 1998 of overburden between existing buildings to a depth of about 2 m. Prior to P&T, groundwater flow was primarily subhorizontal

Conclusions

High-resolution sampling and analysis of bulk CVOC concentrations in rock core delineated TCE and DCE in more detail than possible from water-quality data provided by water samples from typical fractured-rock monitoring wells, leading to refinement of a conceptual site model for CVOC transport in fractured mudstones of the Newark Basin. These data indicated depth- and strata-dependent spatial variability and confirmed the importance of discrete high-permeability bedding-oriented fractures in

Acknowledgments

This study was part of the USGS Toxic Substances Hydrology Program project “Chlorinated Solvents in Fractured Sedimentary Rock” and the Strategic Environmental Research and Development Program project ER-1555. We are grateful for the support of the U.S. Navy and Jeffery M. Dale, Remedial Project Manager. We thank the U.S. Navy's contractors (EA Engineering, Science and Technology Inc.; International Technology Corp.; ECOR Solutions Inc.; Geosyntec Consultants Inc.; Watermark Environmental Inc.)

References (37)

  • R.A. Dearden et al.

    Release of contaminants from a heterogeneously fractured low permeability unit underlying a DNAPL source zone

    J. Contam. Hydrol.

    (2013)
  • P.M. Bradley et al.

    Flowpath independent monitoring of reductive dechlorination potential in a fractured rock aquifer

    Ground Water Monit. Rem.

    (2009)
  • F.H. Chapelle et al.

    Biochemical indicators for the bioavailability of organic carbon in ground water

    Ground Water

    (2009)
  • F.H. Chapelle et al.

    Estimated trichloroethene transformation rates due to naturally occurring biodegradation in a fractured rock aquifer

    Remediation

    (2012)
  • S.W. Chapman et al.

    Combined MODFLOW-FRACTRAN application to assess chlorinated solvent transport and remediation in fractured sedimentary rock

    Remediation

    (2013)
  • G. Cornelissen et al.

    Extensive sorption of organic compounds to black carbon, coal, and kerogen in sediments and soils: mechanisms and consequences for distribution, bioaccumulation, and biodegradation

    Environ. Sci. Technol.

    (2005)
  • M.F. DeFlaun et al.

    Application of bioaugmentation for TCE DNAPL in fractured bedrock

  • D.J. Goode et al.

    Review of aquifer test results for the Lansdale area, Montgomery County, Pennsylvania, 1980–1995

  • D.J. Goode et al.

    Simulation of aquifer tests and groundwater groundwater-flow paths at the local scale in fractured shales and sandstones of the Brunswick Group and Lockatong Formation, Lansdale, Montgomery County, Pennsylvania

  • H. Herpers et al.

    Preliminary report on the geology and ground-water supply of the Newark, New Jersey, area

  • A.D. Hewitt

    Comparison of sample preparation methods for analysis of volatile organic compounds in soil samples. Solvent extraction vs vapor partitioning

    Environ. Sci. Technol.

    (1998)
  • P.J. Lacombe

    Hydrogeologic framework, water levels, and trichloroethylene contamination, Naval Air Warfare Center, West Trenton, New Jersey, 1993–97

  • P.J. Lacombe

    Ground-water levels and potentiometric surfaces, Naval Air Warfare Center, West Trenton, New Jersey, 2000

  • P.J. Lacombe

    Mass of chlorinated volatile organic compounds removed by pump-and-treat, Naval Air Warfare Center, West Trenton, New Jersey, 1996–2010

  • P.J. Lacombe et al.

    Hydrogeologic framework of fractured sedimentary rock, Newark Basin, New Jersey

    Ground Water Monit. Rem.

    (2010)
  • C.A. Lebron et al.

    Dense non aqueous phase liquid (DNAPL) removal from fractured rock using thermal conductive heating (TCH)

  • J.C. Lewis-Brown et al.

    Simulated ground-water flow, Naval Air Warfare Center, West Trenton, New Jersey

  • J.C. Lewis-Brown et al.

    Hydraulic and solute-transport properties and simulated advective transport of contaminated ground water in a fractured-rock aquifer at the Naval Air Warfare Center, West Trenton, New Jersey, 2003

  • Cited by (18)

    • The complex spatial distribution of trichloroethene and the probability of NAPL occurrence in the rock matrix of a mudstone aquifer

      2019, Journal of Contaminant Hydrology
      Citation Excerpt :

      Indicators used to differentiate between the GRY-MAS, GRY-LAM, and BLK-FIS rocks do not always provide a clear distinction, and relatively thin mudstone layers may not be clearly identified. Upon the recovery of rock cores at land surface, samples of the core (approximately 2 cm in length) were separated and processed using methods described in Goode et al. [2014] in preparation for analyses of chloroethene constituents in the rock matrix. In total, 299 core samples were analyzed from the seven coreholes (Table 1) for the concentration of TCE, cis-1,2-dichloroethene (DCE), and vinyl chloride (VC).

    • Variability of organic carbon content and the retention and release of trichloroethene in the rock matrix of a mudstone aquifer

      2018, Journal of Contaminant Hydrology
      Citation Excerpt :

      The magnitude and range of variability in foc in the BLK-FIS rock results in the adsorbed TCE varying as much as an order of magnitude over relative short distances. Investigations of TCE in sedimentary rock conducted by Sterling et al. (2005) and Goode et al. (2014) have demonstrated similar results, with large variability in the TCE content from closely spaced core samples. Because of the large variability in the adsorbed TCE in the rock matrix, point concentrations of adsorbed and aqueous TCE in the rock matrix at a discrete locations may not provide an accurate measure in evaluating the TCE mass in the rock matrix and the longevity of contamination.

    • Characterization of chlorinated solvent contamination in limestone using innovative FLUTe® technologies in combination with other methods in a line of evidence approach

      2016, Journal of Contaminant Hydrology
      Citation Excerpt :

      Characterization of the contaminant distribution in limestone (or chalk) aquifers is especially challenging due to the heterogeneous nature of the limestone, with highly conductive fractures and variable permeability matrix blocks (Witthüser et al., 2003; Loop and White, 2001; Williams et al., 2006). This requires high-resolution data and unbiased samples of the contaminant distribution (as demonstrated by e.g. Parker et al., 2006; Goode et al., 2014). To date most contaminant investigations and monitoring in limestone/chalk and other fractured rock aquifers have been based on water sampling from traditional open boreholes or screened wells.

    View all citing articles on Scopus
    View full text