Hazardous Materials in the Hydrologic Environment: The Role of Research by the U.S. Geological Survey (1996)

Contamination of the environment (surface and subsurface; waters, soils, sediments, and biota) with hazardous materials has occurred through a variety of mechanisms induced by humans. Sources of contamination are classified as either point (a single, concentrated, identifiable source) or nonpoint (a diffuse source). For example, a chemical spill is a point source of contamination, whereas runoff from fertilized farmland is a nonpoint source of contamination. Since the industrial revolution, human activities have served to introduce both anthropogenic and natural materials to the environment in unnatural ways. For example, certain mining operations have produced widespread contamination of surface waters and stream sediments with elevated levels of metals.

A broad spectrum of contaminants have been introduced to the environment in a variety of ways, including surface spills, underground pipeline leaks, surface seepage basins, direct releases to streams or lakes, and underground injection wells. Many industrial operations have resulted in subsurface contamination by solvents. Many facilities that handled petroleum hydrocarbons (tank farms, refineries, pipelines, and gasoline stations) have contaminated the subsurface. Organic contaminants such as solvents and petroleum hydrocarbons have migrated rapidly in the subsurface at these sites, often creating large ground water plumes. Naturally-occurring toxic substances that present human health concerns also have been identified in ground water and surface water. In a number of cases, radioactive materials and trace metals from natural sources have

been found in ground water at levels that exceed public health drinking water standards. For example, LeGrand (1988) reported elevated activities of radium and radon in ground waters of the Piedmont Plateau and the Blue Ridge Mountains. Other toxic substances known to occur naturally at levels exceeding drinking water standards include arsenic, fluoride, lead, strontium, and selenium (Hem, 1992). The USGS maintains and distributes a data base containing chemical analyses of ground water and surface water for many areas of the United States (Hoffman and Buttleman, 1994).

The characterization of sites containing hazardous materials must involve an interdisciplinary approach with personnel with expertise in the fields of hydrology, geology, geochemistry (contaminant distribution), analytical chemistry, microbiology, ecology, statistics, and image processing. Improved understanding of the processes involved in, or affecting, contaminant transport is critical to developing innovative approaches to characterizing both the surface and the subsurface, and ultimately preventing future contamination or remediating sites already contaminated. According to a previous NRC report “the greatest progress will be made if site cleanups are accompanied by investigations aimed at identifying the critical conditions and processes controlling contaminant behavior...” (National Research Council, 1994b). Improvements in process understanding and the development of innovative tools for characterization are needed to advance the state of knowledge of contaminated sites.

In situ remediation represents an attempt to change the physical, chemical, and biological attributes of natural systems to mitigate the adverse effects of hazardous materials in the environment. In order for in situ remediation to be successful, the link between particular attributes of natural systems and the processes affecting the hazardous constituents must be established clearly.

The physical, chemical, and biological attributes of natural systems vary from site to site. Characterizing these properties is important because they determine the response of natural systems to contamination by hazardous materials. The geology and hydrology define the transport characteristics of the system. Dissolved anions and cations, mineral surfaces, and natural organic matter all represent chemical “reagents ” that react in distinctive ways with hazardous waste chemicals. In addition, distribution of bacteria, plants, and other organisms and their level of metabolic activity are affected by the amount of organic carbon sources and the nature and amounts of electron acceptors available.

In recent years significant advances have been made in the characterization of the chemical and biological constituents of natural systems and in the understanding of how they react with hazardous materials. The composition, physical structure, and chemical properties of oxides, clays, and other products of rock weathering have been extensively characterized (Banfield et al., 1991), and have been examined within the context of prevailing hydrologic and biogeochemical conditions (Hem and Lind, 1994; Webster and Jones, 1994). Information of this kind is important for establishing the types of mineral surfaces present in soils and aquifer sediments capable of sorbing pollutant ions (Balistrieri and Chao, 1990; Fuller et al., 1993). For example, iron (II) as a component of silicate and other minerals commonly found in most aquifer materials, has been demonstrated as a strong reducing agent with the capacity to remove many contaminants from the ground water (White, 1990). The efficacy of the iron (II) reduction process has been demonstrated on chromium (VI) in the laboratory (Anderson et al., 1994), and several field-based researchers are examining the efficacy of solids containing both elemental iron and iron (II) as a reactive barrier to remove organic solvents from ground water (Wilson, 1995).

Natural organic matter is an exceedingly complex material that significantly affects fate and transport of pollutants in the environment. Methods have been developed to divide organic matter samples into distinct molecular size and chemical property fractions (Aiken et al., 1992). Chemical derivatization and spectroscopic methods have been used to substantially improve understanding of functional groups and

other structures within natural organic matter (Leenheer et al., 1995). Further research is needed to evaluate sorption of organic pollutants onto soils (Chiou et al., 1983) and to evaluate the role of organic matter in oxidation/reduction reactions, precipitation/dissolution reactions, and complex formation reactions with naturally-occurring and contaminant-derived metals (McKnight and Bencala, 1990).

Bacteria, plants, fungi, and other biological species play an important role in the transformation of both naturally-occurring and contaminant chemicals within the environment. Documenting the distribution and metabolic activity of microorganisms, particularly in subsurface environments, is important for evaluating the potential for biodegradation of contaminants (Vroblesky and Chapelle, 1994; Chapelle et al., 1995). Much has been learned in recent years about the types and abilities of microbial populations indigenous to the subsurface (Thiem et al., 1994). The reduction of selenium (Oremland et al., 1994) and uranium (Lovley and Phillips, 1992) by bacteria has been established recently, providing good evidence that microorganisms play a greater role in the redox transformations of inorganic contaminants than was previously suspected. As additional synthetic organic compounds are shown to biodegrade (Visscher et al., 1994), and as biodegradation in field settings is better understood (Cozzarelli et al., 1994), the need to properly assess the potential for biodegradation becomes more readily apparent. Indeed, a remediation strategy involving no additional active measures is being considered for some contaminated sites where the processes of natural attenuation and biodegradation are acting to remediate the site. This new approach to cleanup of hazardous waste sites is highly dependent on a good characterization of the environment and a sufficient understanding of these processes.

It is important to understand the processes governing natural systems on several temporal and spatial scales. Pertinent temporal scales are linked to a number of factors: the rates of chemical and biological process, the transport of solutes and sediments in the hydrologic cycle, the ecosystem response, and possible human disturbance. Pertinent

spatial scales range from the molecular (where fundamental chemical and biological processes take place) through intermediate scales that govern flow through porous media and the distribution of microorganisms and invertebrates, to scales applicable to human activities and whole ecosystems, to the global scale.

Processes such as precipitation/dissolution, adsorption, complexation, and dispersion, and factors such as oxidation/reduction and pH, control the migration of many constituents in the environment. A better understanding of these processes and controls with a focus on hazardous constituents, will enhance the ability to characterize sites with known contamination and to design effective remediation systems.

Prior environmental research has focused primarily on processes occurring in a single medium: air, soil, or water. Natural environments are open systems, however, and processes acting across media are of fundamental importance. In order to understand the dynamic behavior of natural systems, an interdisciplinary approach is required.

Significant advances in knowledge of subsurface contaminant migration processes have occurred in the last thirty years. Improvements in understanding of processes such as facilitated transport, adsorption, and dispersion have enhanced contaminant distribution prediction and remediation systems design. Progress also has been made in understanding the geological processes that formed most subsurface units. In particular, more has been learned about depositional models, diagenetic processes that alter geologic materials, subsurface heterogeneities, and the hydrology of ground waters that flow through complex subsurface matrices. Additional advances are needed to achieve a better understanding of the heterogeneous nature of the subsurface, however.

Significant progress has been made in the last 20 years in understanding the ecology of subsurface microorganisms and the role they play in the fate and mobility of contaminants. Microorganisms have been shown to transform hazardous materials to products that are either harmless or less hazardous, to convert them to forms with differing solubility, or to sorb them onto cell surfaces. In addition, notable advances have been

made in knowledge of the degradability of numerous hazardous compounds, including the identification of specific degradation pathways. Many questions with respect to biological processes in the subsurface, such as microbial transport (Hurst, 1991), and the fate and degradability of contaminants, still remain unanswered, however.

Because contaminants such as gasoline and solvents have entered the subsurface as separate phases (i.e., not as a dissolved phase in water) it is also necessary to understand multiphase flow in the subsurface. Most research in this area has concentrated on model studies and well-controlled laboratory investigations. This theoretical work, originally developed within the petroleum industry, provides an important theoretical and methodological framework for subsequent work that has occurred in the field of hydrogeology. The problems are even more complex in the field of contaminant transport because they involve interphase mass transfers. Compositional models that incorporate interphase transfer (Abriola and Pinder, 1985a,b; Baehr and Corapciouglu, 1987) have been used as the principal approach to modeling nonaqueous phase liquid (NAPL) flow. True multiphase capabilities incorporating complex patterns of gas flow and mass removal have been developed to support theoretical investigation of remedial approaches such as gas sparging or soil venting. Mass transfer between NAPLs and water and between aqueous and gas phases is being studied to improve the knowledge of contaminant migration through the subsurface (Anderson et al, 1992; Miller et al., 1990; Whelan et al., 1994).

There is a deficiency of fundamental field and laboratory data concerning multiphase flow parameters relevant to contaminant systems. Physical models have been utilized to improve the state of understanding. Mass transfer reactions typically encompass families of nuclear, chemical, and biological processes. Although some of these processes like radioactive decay are well understood, those processes involving multiple chemical species and biological reactions are much less understood and provide a major focal point of contemporary research in contaminant hydrogeology. Valid conceptual and mathematical representations exist for many of these processes, but they have not yet been applied to solving real world problems. For example, the use of overly simplistic models such as distribution coefficients to describe sorption is now being re-evaluated in terms of better conceptual and mathematical models (Barber,

1994; Harvey et al., 1989; Harvey and Garabedian, 1991; Stollenwerk, 1991). Other important areas of research include the modeling of complex species in solution (MINTEQ: Felmy et al, 1983; EQ3/EQ6: Wolery, 1979) and the incorporation of inorganic reactions into flow and mass transport models (Cederberg, 1985; Liu and Narasimhan, 1989; Narasimhan et al., 1986).

The diversity and complexity of subsurface systems require an understanding of coupled flow processes. Coupling of thermal, hydrologic, mechanical, biological, and chemical processes is required to obtain a complete understanding of the subsurface. Tsang (1987) provided a general overview of the commonly studied coupled problems in hydrology. The most advanced studies to date involve codes such as V-TOUGH, developed for the Yucca Mountain nuclear waste repository program to predict the response of the hydrologic system to significant repository heating (Buscheck and Nitao, 1992). Other studies have examined density-driven transport of dense hydrocarbon vapors in partially saturated media (Mendoza and Frind, 1990a,b) and the development of instabilities in variable density flow (Schincariol and Schwartz, 1993).

The study of fractured media has been a major focus of a group of researchers over the last 30 years. New knowledge about how fluids move in the subsurface through fractured media has been obtained through advancements in fracture flow modeling and field experiments. For example, the USGS has recently conducted in-depth, multidisciplinary studies of site characterization and ground water movement in fractured rocks at the Mirror Lake site in New Hampshire (Hsieh et al., 1993). These studies have brought together hydrogeologists, geophysicists, geochemists, structural geologists, and numerical modelers to address fundamental questions of fluid flow in such environments. Theoretical work in this field has continued to progress through the development of more realistic fracture flow codes. The state-of-the-art in discrete fracture models is represented by codes such as NAPSAC (UK Harwell) and FracMan/MAFIC (Golder Associates, 1988). Recent work (Sudicky and McLaren, 1992) has extended the discrete modeling approach to accommodate complex fracture matrix coupling in both flow and contaminant transport. Most field and laboratory studies related to fractured rock problems continue to be motivated by the need to assess the implication of fracturing in relation to waste storage and contaminant transport.

When potentially hazardous materials are accidentally spilled or are deliberately applied on vegetation or soils, they may be transported to surface waters by direct surface runoff, or to aquifers by water percolating through the unsaturated zone. Surface waters deserve a great deal of attention because of their widespread use as drinking water sources, their importance to fisheries and as habitats for wildlife, and their role in the hydrologic cycle. Energy production, manufacturing, agricultural production, mining, and waste treatment all are performed in close proximity to surface waters, and all represent potential sources of contamination. In addition, surface waters receive both ground water and atmospheric inputs of contaminants. Sediments, which often contain considerable contaminant levels as a result of past activities, are both contaminant sources and sinks with respect to the water column.

Intensive field and analytical research on the fate and transport of surface applied chemicals has been carried out since the mid-1960s. Much of this research has been done by the Agricultural Research Service (ARS) of the U.S. Department of Agriculture (USDA) and by universities. These groups have developed several computer models that predict the fate and transport of agricultural chemicals in order to assist policy-makers in making regulatory and policy decisions (U.S. Department of Agriculture, 1980; Smith, 1992). However, there are many uncertainties involved in predicting the fate and transport of surface applied chemicals because of the complicated nature of the processes and the unpredictability of factors such as precipitation (Woolhiser, 1976).

The flow and chemical composition of upland streams and rivers are strongly linked to hydrologic processes and biogeochemical processes within the watershed. There is a clear need to evaluate the causative factors in temporal variability, especially as they pertain to the movement of hazardous materials. McKnight and Bencala (1990) and McKnight et al. (1992) of the USGS have made significant progress in understanding the temporal behavior of iron, aluminum, and natural organic matter in streams and rivers, and their biogeochemical linkages. In these studies, the exchange of water and solutes with adjoining sediments and aquifers has been found to be significant.

Lowland rivers, lakes, and estuaries are linked to larger and more complex watersheds. Establishment of sources and sinks to such systems is essential, as is the development of a hydrologic model that can account for mass transport. The dynamic behavior of pesticides and other organic contaminants in San Francisco Bay has been found to be a strong function of the distribution between water and suspended sediment (Domagalski and Kuivila, 1993). Distinguishing hydrologic inputs from atmospheric inputs is important for evaluating the efficacy of existing regulatory controls on contaminant release. Estimates of atmospheric inputs into the Chesapeake Bay, for example, have been revised upward in recent years (Baker et al., 1994).

When dealing with sites contaminated at the ground surface, ecological characterization is essential. Improving the understanding of stresses and changes that have occurred as a result of contamination at a site is critical to the long-term goal of site restoration. Ecological processes must be understood so that contaminant migration and processes of natural attenuation can be better understood.

Characterization methods available to the geosciences community have improved vastly in recent years. Some new methods represent cost-effective and reliable alternatives to existing characterization methods, whereas others provide insight into parameters and processes that was not available in the past.

In order to obtain information about the geology of a site, samples are generally collected by drilling holes into the subsurface. Indirect information about the subsurface geology also can be obtained using both surface and downhole geophysical tools. Existing methods for characterizing the hydrology of a site include the installation of wells and piezometers from which aquifer tests and tracer tests can be performed. In order to obtain information about the geochemistry of a site (contaminant distribution), samples are collected from soils (both at the surface and downhole), from ground water, and from vapor samples in the unsaturated zone.

Innovations in characterization technology have been developed to promote collection of better data at lower cost. Emphasis has been placed on innovative drilling technologies that minimize the amount of hazardous waste extracted while maintaining the integrity of samples collected. New techniques to perform chemical analyses in the field have been developed to lower overall analytical costs and allow for real-time data collection. Innovative geophysical tools have been developed recently to enable collection of better data about the geology, geochemistry, and hydrology of the subsurface, while also minimizing the amount of intrusion into contaminated zones and reducing costs. Chemical isotopes have been utilized to characterize the age of ground waters and to trace or delineate ground water flow paths. All this information is used to improve the understanding of the migration of contaminants with the ultimate goal of promoting better design of remediation systems.

Innovative drilling technologies recently applied to hazardous waste site applications include the cone penetrometer, which collects geologic and geochemical data before new drilling locations are selected. These data are used to design a more standard drilling program for characterization of the site and for installation of monitoring wells. Other innovative drilling technologies include sonic drilling, which minimizes hazardous waste materials extracted during drilling while maintaining the quality of the core that can be obtained from the subsurface (Barrow, 1994). Horizontal drilling has been modified for drilling at hazardous waste sites, enhancing access to the subsurface and promoting characterization and remediation of sites otherwise inaccessible, such as under large buildings or under landfills (Kaback et al., 1989).

Innovative sampling technologies include the SEAMIST™ liner that collects vapor samples at discrete depths in the unsaturated zone. The HydroPunch™ and the BAT™ sampler have been used to collect depth-discrete ground water samples from a single borehole (Kaback et al., 1990). Fiber optic sensors have been developed to detect subsurface contamination in monitoring wells and have been adapted to the cone penetrometer to provide real-time data in the field (Colston et al., 1992). Innovations in three-dimensional image analysis have allowed scientists to create better visualizations of subsurface contamination (Eddy and Looney, 1993). Geophysical methods for imaging the subsurface also have been improved. For example, crosshole tomography using electrical

resistivity, electromagnetics, and seismic sources has been demonstrated as a tool to image subsurface geology, hydrology, and the effects of in situ remediation on the subsurface.

Fate and transport modeling has been advanced using new codes developed to mimic real subsurface conditions, such as heterogeneity, fractured bedrock, and adsorption. The understanding of subsurface microorganisms, their diversity, and their innate ability to remediate contaminants in the subsurface has developed recently into an important research area. Innovative techniques to characterize these intrinsic inhabitants of the subsurface using DNA probes are an example of recent advances in the state-of-the-art (Hazen and Jimenez, 1988).

The quality and utility of analytical results depend upon many elements of the research protocol. Samples must be collected over spatial and temporal scales that capture systematic and stochastic variations in the parameters under study. Numerous precautions must be taken to maintain the integrity of samples and minimize contamination. Because each analytical technique has its strengths and limitations, it is important to perform several complementary techniques on the same field samples.

Surface water contamination by toxic elements is often demonstrated by comparing total concentrations in waters receiving anthropogenic inputs to concentrations from upstream, pristine waters. Ultraclean sampling, handling, and analysis and careful comparison with reagent and instrument blanks are necessary to obtain reliable trace metal data from surface waters. Failure to follow these procedures can yield estimates of toxic metal concentrations that are two orders-of-magnitude too high (Benoit, 1994). In combination with ultraclean techniques, resolution of naturally-occurring isotopes (e.g. Erel et al., 1991) and rare earth element profiles (Olmez et al., 1991) can be used to resolve anthropogenic inputs from those derived from natural sources.

Characterization of natural organic matter is needed to evaluate its ability to form complexes with toxic metals, form covalent compounds with pesticides, participate in dissolution/precipitation reactions of minerals, serve as a carbon source for bacteria, and form chlorination byproducts during drinking water chlorination (Aiken and Cotsaris, 1995). USGS scientists have been at the forefront of developing new and innovative means of collecting (Aiken et al., 1992) and characterizing (Leenheer et al. 1995) natural organic matter, and have developed a comprehensive

understanding of its role in natural biogeochemical processes (Averett et al., 1989). In addition, Nitrogen-15 and carbon-13 Nuclear Magnetic Resonance (NMR) methods to obtain fundamental information regarding natural organic matter functional groups are currently under development (Thorn et al., 1992). Capillary electrophoresis, which separates organic molecules according to molecular charge and hydrodynamic radius, has been shown to resolve natural organic matter sub-fractions (Garrison et al., 1995). In combination with established detection methods, capillary electrophoresis should be an effective new tool for characterizing natural organic matter.

Gas chromatography, in combination with electron capture detection or with mass spectrometry, represents one of the most powerful analytical techniques available today. Neutral, hydrophobic organic contaminants are most amenable to such analysis. In the case of polychlorinated biphenyls (PCBs), resolution of individual congeners at environmentally-relevant levels is now possible (Eganhouse and Gossett, 1991). Changes in congener profiles in space and in time are now widely used to explore the physical, chemical, and biological processes acting upon them. Ionized organic contaminants frequently require derivatization and sample enrichment. The determination of anionic surfactants in sewage effluent and ground water samples has, however, been effectively demonstrated (Field et al., 1992).

Better understanding of the processes that affect contaminant migration are critical to cleaning up the nation's hazardous waste sites. Good characterization data are required to design effective and efficient remediation systems. Improved understanding of processes such as adsorption, desorption, facilitated transport, and sediment-water interactions are required so that superior systems for removing or immobilizing contaminants can be designed and developed. Characterization of sites with separate phase contaminants such as light nonaqueous phase liquids (LNAPLs) and dense nonaqueous phase liquids (DNAPLs) is critical to remediation of sites contaminated with organics. The principles developed within the petroleum industry regarding petroleum migration and

extraction from the subsurface have been, and should be applied to studies of contaminant fate and transport and remediation designs.

Improvements in the understanding of subsurface heterogeneities also will be required for better predictions of contaminant fate and transport. A major research need with respect to characterizing biological processes in the subsurface is to improve the fundamental understanding of microbial populations in context with the heterogeneous physical and chemical conditions of the subsurface. Questions remain as to how microbial populations develop and are maintained in aquifers through periods of environmental stress caused by insufficient substrates, nutrients, moisture, or other undesirable conditions. Associated with these questions are the issues of microbial transport and survival within saturated and unsaturated, aerobic and anaerobic zones, as well as the issue of relating microbial populations to flow patterns within an aquifer.

More tools are needed to characterize the subsurface non-invasively or with minimal invasion. Improved field screening of contaminants could save millions of dollars. Development of tools to characterize sites with mixtures of contaminants will be necessary as more is learned about the nature of contaminant problems. Future research must integrate the understanding of surface and subsurface processes and closely examine the interaction between ground water and vadose zone processes. Hazardous waste systems must be examined as a whole by integrating concepts from a variety of fields.

The USGS has made significant contributions to the understanding of the hydrologic cycle and the subsurface in both pristine and contaminated settings. This research has concentrated on improved understanding of processes rather than the creation of new technologies. Future opportunities within this arena will likely be extensions of previous or existing work, rather than totally new endeavors. Process-related research topics of broad scope that require the multidisciplinary resources and expertise available within the USGS are listed below:

• Hydrological, biological and geochemical processes within fractured and heterogeneous media.

• The effects of physical form or phase (e.g., NAPLs, gases, sorbed) on the availability of contaminants to geochemical and microbial actions and reactions.

• Incorporation of biological processes into contaminant fate modeling.

• Microbial transport through the subsurface.

• Microbiological adaptation and contaminant degradation under conditions of typical subsurface physical and chemical stresses.

• Development of analytical tools to evaluate chemical, hydrological, and biological processes directly within the subsurface environment.

• Improved prediction capability for heterogeneities in the subsurface.

• The effect of nutrient levels (nitrogen and phosphorus) on the chemistry, biology, and transport of contaminants in surface waters.

• Sediment-water interactions, such as contaminant entrainment in sediments by particle settling, resuspension by storms, and release caused by chemical and biological processes within the sediment.

• Direct photolysis of contaminants in surface waters, and the reaction of contaminants with reactive species generated through the photolysis of organic matter and other natural solutes.

• The production and decomposition of natural organic matter.