Few comprehensive studies of contamination of karst by chlorinated solvents have been published (Barner and Uhlman, 1995; Cary McConnell, University of Missouri, written commun., 1996; Robert Westly, SCS Engineers, written commun., 1996). Most articles relevant to this topic fall into the following categories:
Published case studies of chlorinated-solvent contamination in karst are few. None meet current standards for definitive characterization of contaminated karst sites, nor the more stringent standards applicable to dense nonaqueous-phase liquid (DNAPL) contamination at such sites (Field, 1988a; Quinlan, 1989, 1994; Barner and Uhlman, 1995; Quinlan and others, 1995). A few case studies of DNAPL contamination in fractured rock are relevant to the potential for DNAPL movement in carbonate rocks with little dissolutional enlargement of fractures.
Chlorinated solvents present unique contamination problems when released onto or into the ground. A number of reports provide comprehensive descriptions of chlorinated-solvent contamination (Schwille, 1988; Mercer and Cohen, 1990; Wilson and others, 1990; Cohen and Mercer, 1993; Pankow and Cherry, 1996). A few of these reports discuss the movement and dissolution of DNAPL in fractured rock, but none treat karst as a separate type of setting. Several other reports discuss volatilization of DNAPL (Marrin and Thompson, 1987; Nitao and others, 1991), DNAPL movement in porous media (Kueper and others, 1989, 1993; Poulsen and Kueper, 1992; Michalski and others, 1995; Tuck and Iversen, 1995) and in fractures (Helton, 1987; Cherry, 1989; Quigley and Fernandez, 1989; Kueper and McWhorter, 1991; Ford, 1993), the dissolution of DNAPL pools (Johnson, 1987; Feenstra, 1990; Anderson and others, 1992; Johnson and Pankow, 1992; Uhlman, 1992), and the dissolution and diffusion of DNAPL from cracks into the rock or clay matrix (Vogel and Giesel, 1989; Watts and Cooper, 1989; Parker and others, 1994; Parker and Cherry, 1995; VanderKwaak and Sudicky, 1996).
Many previous studies of karst exist. Comprehensive works include general texts on karst (Herak and Stringfield, 1972; Sweeting, 1972; Milanovic, 1981; White, 1988; Ford and Williams, 1989; Higgins and Coates, 1990) and discussions of carbonates, dissolution, and conduits (Bebout and others, 1979; Brahana and Hollyday, 1988; Brahana and others, 1988; White and White, 1989). Other studies particularly relevant to chlorinated-solvent contamination focus on properties of karst that control DNAPL movement and on the potential for contamination of karst ground water.
Karst soils and subsoils typically have sufficient secondary porosity to allow infiltration of low-viscosity DNAPL (Thomas and Phillips, 1979; Watson and Luxmoore, 1986; Quinlan and Aley, 1987; Wilson and others, 1991a, 1991b). Other studies have reached the same conclusion for clays and clayey soils in general (Cherry, 1989; Everts and others, 1989; Wells and Krothe, 1989).
Sinkhole development creates opportunities for the direct introduction of contaminants into karst aquifers, and is controlled by a variety of geologic factors (Kemmerly, 1980, 1981; Magdalene and Alexander, 1995; White and White, 1995). Human activities, including the disposal of storm water and liquid and solid wastes, can contribute to karst subrosion and sinkhole development (Royster, 1984; Newton, 1987; Field, 1993).
Epikarst, the hydraulically important transition zone including the base of the regolith and the irregular and dissected bedrock surface, is the focus of only a few articles (Julian and Young, 1995). Epikarst is discussed (with or without the use of the term "epikarst") in several more general reports (Field, 1988a, 1989, 1993; Ford and others, 1988; White, 1988; Quinlan and others, 1992) and in a few site studies (Chieruzzi and others, 1995; Glover and others, 1995).
Karst is more vulnerable to ground-water contamination than are many other hydrogeologic settings (Quinlan and others, 1992; Fetter, 1993; Field, 1993). Adequate characterization of contaminant movement in karst is difficult, and the use of approaches well-suited for other settings may produce ambiguous or misleading results (Field, 1988b; Alexander, 1989; Rubin, 1992; Quinlan, 1994; Barner and Uhlman, 1995). Dense immiscible contaminants may become isolated in or under sediment deposits commonly found in conduit systems (Gale, 1984; Palmer, 1986). The movement of the dissolved contaminant can be complex and unpredictable (Becher, 1989; Seiler and others, 1989; Sudicky and MacQuarrie, 1989; Jancin and Ewart, 1995). Various other authors have investigated specific cases of contamination (Harvey and Skelton, 1968; Alexander, 1989; Hannah and others, 1989; Waite and Thomson, 1993; Currens, 1995; Sasowsky and others, 1995).
Numerical models of multiphase flow have been developed for application to porous media in the laboratory setting (Guarnaccia and Pinder, 1992; Zhan, 1992; Al-Sheriadeh and Illangasekare, 1993), and have been applied also to karst field sites (Faust, 1985; Guswa, 1985; Faust and others, 1989). The difficulties of modeling aqueous flow in karst are well known (Quinlan and others, 1995). The necessity to model multiple-phase, density-driven flow greatly complicates the modeling task, particularly in extremely heterogenous, anisotropic karst settings. The extent to which digital modeling adds to the understanding of chlorinated-solvent contamination is debatable (Schmelling and Ross, 1989; Huling and Weaver, 1991; Cary McConnell, University of Missouri, written commun., 1996).
The movement of DNAPL's in karst landscapes has been studied in several locations, but only a few of these investigations have resulted in published reports (Cary McConnell, University of Missouri, written commun., 1996; Robert Westly, SCS Engineers, written commun., 1996). Near Ville Mercier, Quebec, chlorinated solvents penetrated 30 meters (m) of sandy glaciofluvial deposits, bypassed 3 m of clayey till, and entered the fractured sandy dolomite underlying the site (Martel, 1988). At the Bear Creek Burial Grounds at Oak Ridge National Laboratories in Tennessee, investigators reported DNAPL moving down the dip of steeply dipping shale and limestone (Doll, 1992; Kueper and others, 1992; Shevenell and others, 1992). At Pinto, West Virginia, TCE descended through the alluvium of the North Branch Potomac River and deep into karst bedrock, making complete remediation infeasible (Ford, 1993). Chieruzzi and others (1995) used well tests to demonstrate the dominance of diffuse over conduit flow in a southwestern Kentucky karst setting contaminated with DNAPL.
Studies have been performed at several sites above the Niagara Escarpment in and near Niagara Falls. Mercer and others (1983) studied the Love Canal landfill at Niagara Falls, New York. At the S-area landfill in Niagara Falls, the glaciolacustrine clay confining layer over the dolomite prevented descent of DNAPL; at the Hyde Park landfill, DNAPL is present near the base of the dolomite aquifer (Faust and others, 1989). McIelwain and others (1989) characterized a site near Smithville, Ontario, where a solution of oil, polychlorinated biphenyls, chlorinated solvents, and other organic compounds penetrated through 5 to 10 m of weathered glaciolacustrine silty clay, then moved downward along joints and horizontally along bedding planes through limestones and dolostones of the Niagara Escarpment.
Studies of DNAPL contamination at sites underlain by fractured noncarbonate rocks demonstrate that many such rocks contain sufficient small openings to allow penetration by DNAPL (Kraus and Dunn, 1983; Nichols and Gibbons, 1988; Holmes and Campbell, 1990; Kueper and others, 1992; U.S. Environmental Protection Agency, 1992a; National Research Council, 1994). These results are relevant to carbonate rocks in which little dissolutional enlargement has occurred. In most cases, field data show that DNAPL descended until the fracture system pinched out. In other cases, confining units stopped or deflected DNAPL movement (Becher, 1989; Lindhult and others, 1990).
|Next||Title Page||Table of Contents||List of Illustrations||List of Tables||Conversion Factors and Vertical Datum||Glossary|