5.1. General overview
Research on the biogeochemical and microbial processes that underlie the success of natural attenuation, biostimulation, and/or bioaugmentation remedial strategies for contaminant organics in groundwater has received much attenuation in recent years (Scow and Hicks, 2005). Both natural and intentional biodegradation and imbibition of organics can occur in con- taminated subsurface environments as long as available electron donor and acceptors, and nutrients are available, and if subsurface geochemical and hydrologic conditions are conducive to microbial growth and activity (Aulenta et al., 2006; Xie et al., 2006). Understanding site hydrology is critical to the evaluation of contaminant plume development since varia- tions in terminal electron accepting processes (TEAPs) and microbial physiological processes occur along groundwater flow paths and with depth within contaminant plumes (Chapelle, 1993; Chapelle et al., 1995;
Lovley, 2001; Smith, 1997). Characterization of coupled physicochemical and hydrologic processes is important because the degradative potential of specific compounds within a plume varies depending on the TEAP (Fig. 11). TEAP zones are often delineated by measuring concentrations of electron donors and acceptors in groundwater samples; however, inter- pretation of this data is difficult because these concentrations are influenced by not only microbial reactions, but by hydrological processes such as advection, dispersion, diffusion, and coupled geochemical reactions. Plumes of contaminated groundwater often exhibit sharp vertical or lateral gradients in contaminant concentration, aquifer geochemistry, and redox conditions (Bennett et al., 2000; Christensen et al., 2000; NRC, 2000), thereby providing opportunities for fundamental investigations of microbial com- munity diversity, ecology, and function within TEAPs. Chemical gradients are also used to infer natural attenuation processes in contaminated aquifers (NRC, 2000) where TEAP shifts are typically accompanied by changes in aquifer geochemistry (Bennett et al., 2000) and chemical shifts due to biodegradation reactions (NRC, 2000).
Haack and Bekins (2000)discussed the prevalence of various microbial populations in organic-based contaminant plumes and they focus on non- manipulated systems and experiments that have been conducted in situ or
48 Philip M. Jardine
under conditions representative of a contaminated field site. They suggested that the effective biodegradation of subsurface organic contaminants requires that the appropriate microbial populations be present and that favorable subsurface geochemical and hydrological conditions prevail.
They stress that interdisciplinary approaches to site investigations have significantly improved our understanding of hydrogeochemical and micro- biological interactions within these systems. Culture-based and molecular analyses of microbial populations in subsurface contaminant plumes have revealed significant adaptation of microbial populations to plume environ- mental conditions where spatial and temporal variability of geochemical and hydrological conditions can significantly influence subsurface microbial community structure and thus the rates and mechanisms of biodegradative processes (Krumholzet al., 1996). Such variations in physical and chemical properties of an aquifer and the aqueous concentration of contaminants and other important reactants in turn significantly impact the spatial distribution of microbial communities in aquifers (NRC, 1993). For example, the presence of macroscale depositional layers of different permeability, solution
O
SO42− S2−
SO42−
CO2 O
N2 CO2
O
O2− H2O CO2 O
Fe(III) Fe(II) CO2
O CO2
CH4 Contaminant
source
Methanogenic
Direction of groundwater flow
Unsaturated zone Groundwater
Sulfate
reducing Aerobic
Nitrate reducing Fe(III)-reducing
NO−3 NO−3
O2 O2
Figure 11 Schematic illustration of changes in terminal electron acceptor processes (TEAPs) along an organic contaminant plume where highly reducing conditions prevail near the source and oxic conditions exist at the plume fringe (fromLovley, 2001).
Influence of Coupled Processes on Contaminant Fate 49
channels, and fractures zones, superimposed on microscale media hetero- geneities add to the irregularities of subsurface formations and the spatial distribution of microbial habitats (Atlas and Bartha, 1987; Paul and Clark, 1989). As well, microorganisms with specific degradative capacities may only be found in favorable habitats that exist only in a portion of a contaminant plume (Anderson et al., 1998). Thus, understanding spatial and temporal changes in hydrology and chemistry, as related to microbial capabilities over time will better facilitate forecasts of the long-term fate of contaminant plumes.
Variability in the physical structure of aquifer materials also influences bacterial transport (Harvey et al., 1993). Lawrence and Hendry (1996) reviewed the role of geologic, hydrologic, and solute variables on bacterial transport through geologic media. They concluded that variations in site hydrogeology may significantly influence the location and community structure of subsurface microbes through differential transport mechanisms influencing bacterial movement.Ulrich et al. (1998)used microscale radi- ography to show that sulfate-reducing activity was significantly less in clayey sediment layers where water flow was restricted versus more unconsolidated sediment layers where advective flow dominated. However, Ludvigsen et al. (1999) used PFLA analyses to show increased microbial populations and changes in community structure within fine-grained clay and silt layers relative to coarse-grained sand layers in an aquifer contaminated with landfill leachate.
Baker and Herson (1990) discussed the use of various manipulative bioremediation strategies for degrading organic hazardous waste in subsur- face environments. They suggested that these technologies depend upon the alteration of the physical and chemical conditions in the subsurface envi- ronment in order to optimize microbial activity. They reviewed the micro- biological, hydrological, and geochemical factors that should be considered in evaluating the appropriateness of bioremediation of hazardous organic waste-contaminated aquifers and subsurface soils. They stress that enhanc- ing biodegradation or biostimulation at contaminated sites requires altera- tions on the site’s physical and chemical characteristics in order to increase biodegradation by indigenous organisms. Perturbation of environmental factors such as pH, temperature, nutrient availability, and physical structure on microbial abundance, diversity, activity, and function are all important considerations for the successful application of bioremediation at organic contaminated sites.
Recently, Scow and Hicks (2005) recommend that three lines of evidence be used to assess the effectiveness of organic contaminant biore- mediation: (1) a decrease in contaminant mass and concentration as a function of time, (2) observed changes in hydrological and/or geochemical processes thus providing indirect evidence, and (3) in situ or microcosm studies providing direct evidence for biodegradation (NRC, 2000; Smets
50 Philip M. Jardine
and Pritchard, 2003). The authors also suggest that stable isotope tech- nologies and various microbial molecular tools have been employed to quantify remedial success and the organisms responsible for contaminant transformation and destruction within organic plumes.
In this section, the influence of coupled processes on subsurface con- taminant organic fate and transport are discussed. The section is divided into specific organic contaminant types with a focus on recent field-scale rele- vant examples. The section ends with a discussion of recent modeling strategies that incorporate coupled processes in the simulation of the fate and transport of nonaqueous phase organic constituents in subsurface environments.
5.2. Chlorinated solvents
Nonaqueous phase liquids (NAPLs) that have a density greater than water and which tend to sink in subsurface environments are known as dense nonaqueous phase liquids (DNAPLs) with chlorinated solvents, such as trichloroethylene (TCE) and perchloroethylene (PCE), being some of the most common DNAPL constituents. The chlorinated solvents PCE, TCE, and other highly chlorinated aliphatic hydrocarbons (CAHs) such as 1,1,2- trichloroethane (TCA) and chloroform (CF) are widely used in various industrial processes, mainly as solvents in dry-cleaning and semiconductor manufacture. They are highly toxic and are known or suspected carcinogens.
Chlorinated solvents can be transformed into a variety of products within subsurface environments via biotic and abiotic processes (Aulenta et al., 2006; Lenczewski et al., 2003; Lorah and Voytek, 2004; Vogel et al., 1987). Most CAHs do not support microbial growth under aerobic condi- tions; however, many are anaerobically transformed to less chlorinated or nonchlorinated compounds through dechlorination reactions. Nonchlori- nated products are typically harmless and many of the less chlorinated daughter products of dechlorination are biodegraded under aerobic condi- tions as well. Two distinct microbial-mediated reductive dechlorination reactions exist: hydrogenolysis and dichloroelimination (Aulenta et al., 2006) where hydrogenolysis (i.e., reductive dechlorination) involves the replacement of chorine with hydrogen with a net input of one proton and two electrons from an external donor; whereas, dichloroelimination results in the replacement of the chlorine substitutes and the formation of a double bond between the two carbon atoms, with an input of two electrons from an external donor. Most known bacterial isolates that utilize CAHs as terminal electron acceptors require H2as an electron donor, and numerous organic fermentable substrates such as lactate, glucose, ethanol, and so on can be used as a source of H2for enhancingin situdechlorination. A drawback of using these substrates is the possible accumulation of fermentation products and the
Influence of Coupled Processes on Contaminant Fate 51
generation of large amounts of biomass in the subsurface that can modify groundwater hydrologic flow paths and induce porosity clogging.
As mentioned above, PCE and TCE form DNAPL that sinks through permeable soil or aquifer material until a nonpermeable or restrictive layer (e.g., nonuniform soil texture) occurs resulting in the formation of DNAPL pools (Fig. 12). Some of the DNAPL may volatilize to the atmosphere or become trapped in soil pores, while other portions of the DNAPL may be sorbed on soil particles or partitioned into organic matter (Bollaget al., 1980;
Schwarzenbach and Giger, 1985). During DNAPL transport, hysteretic capillary forces result in a complex pattern of discontinuous globules of organic liquid in the pore structure of the subsurface. Entrapped DNAPL mass dissolves slowly into flowing soil and groundwater and thereby acts as a long-term source. Because of this, remedial technologies have used alcohol and surfactant flushing to enhance the recovery of the contaminant organic (Christet al., 2005; Londerganet al., 2001; Ramsburget al., 2004). As might be expected, effective dechlorination by microorganisms acts to drive the dissolution of DNAPL by reversing the gradient that causes DNAPL forma- tion (Yang and McCarty, 2000). This microbial-enhanced DNAPL dissolu- tion is effective as long as toxic byproducts, such as vinylchloride (VC), do not accumulate in the subsurface. The most effective means of DNAPL removal from the subsurface is to take advantage of coupled processes where aggressive physicochemical processes are used to remove the bulk of the solvent mass and bioremediation is used as a polishing step to the
DNAPL residual
Bedrock Vadose zone
Capillary fringe Water table
Ground water flow direction
After NRC, 1994
DNAPL as separate fluid phase Dissolved DNAPL in ground water Vapors emanating from DNAPL Clay layer
Figure 12 Schematic illustration of DNAPL transport through the vadose zone and entry into the groundwater. As DNAPL migrates downward through the vadose zone it leaves residual saturation behind, until it reaches the water table. Pooling occurs on restrictive layers (e.g., clay layers) causing lateral spreading (from http://pangea.
stanford.edu/research/hydro/).
52 Philip M. Jardine
contaminant mass flux emanating from the remaining DNAPL. Successful field-scale bioremediation of chlorinated solvents requires careful design, operation, and monitoring of electron donor additions, hydrological and geochemical conditions, and the propensity for contaminant reduction (Morseet al., 1998).
Lenczewskiet al. (2003)investigated the influence of coupled processes on the fate and transport of chlorinated solvents in a fractured shale ground- water environment on the Oak Ridge Reservation in eastern Tennessee, USA. The field study offered an excellent example of how natural microbial attenuation processes transformed TCE into a variety of endproducts depending on the spatial and temporal changes in hydrology and geochem- istry. The field site consisted of a 50-m long transect of multilevel ground- water sampling wells that followed strike parallel geology and extended from numerous waste burial trenches to a seep exiting into a perennial stream. The wells were situated within either a fast flowing fracture regime or a slow flowing matrix regime and the spatial and temporal variability of TCE, 1,2-DCE, VC, ethylene, ethane, methane, and various geochemical redox indicators were quantified as a function of time. The concentration of organics followed the general trend 1,2-DCE>ethylene>VCTCE with TCE concentrations being 10-fold higher in the waste trenches relative to downgradient sampling wells, whereas VC, 1,2-DCE, and ethylene con- centrations in the waste trenches were similar or slightly higher than the downgradient groundwater monitoring wells. TCE concentrations disap- peared within 10 m from the end of the waste trenches and 1,2-DCE disappeared just prior to the seep, 50 m downgradient the trench source.
VC and ethylene were still present at the seep, with ethylene showing peak concentrations at this locale. These results indicated that anaerobic reduc- tion of the chlorinated organics was occurring. In addition, the presence of high concentrations of methane throughout the site was also an indication of anaerobic metabolism (Lenczewskiet al., 2003). The accumulation of VC in many locations was consistent with the notion that conversion of VC to ethane is usually the rate limiting step during reductive dechlorination of chlorinated solvents (Ballapragadaet al., 1997).
Another interesting finding at the site was that the concentration of chlorinated organics was slightly higher in the matrix relative to the fracture regime, whereas the concentration of ethylene was lower in the matrix relative to the fracture regime. These results suggest that the more hydro- logically active fracture regime was slightly more effective in the anaerobic reduction of the chlorinated organics. Temporal variability of TCE and its degradation products was slight, with a general increasing concentration trend of chlorinated organics and dissolved gases as the site hydraulic gradient increased (a response to increased storm events during the winter and spring months). This may imply that the intrinsic bioremediation scenario at the site was less effective at higher discharge rates. Measurements
Influence of Coupled Processes on Contaminant Fate 53
of redox potential at the site indicated that iron-reduction, sulfate reduction, and potentially methanogenesis were occurring and are conducive to dechlorination of TCE. Bacterial enrichments of groundwater samples revealed the presence of methanotrophs, methanogens, iron-reducing bac- teria, and SRB, all of which have previously been implicated in anaerobic biodegradation of TCE. 16S ribosomal DNA (rDNA) sequences from DNA extracted from groundwater were similar to sequences of organisms previously implicated in the anaerobic biodegradation of chlorinated sol- vents. The combined data strongly suggest that anaerobic dechlorination of TCE to VC and ethylene was occurring (Lenczewski et al., 2003). The presence of methane oxidizers (methanotrophs) also suggested possible zones of oxygenated groundwater, which was confirmed with on-site DO measurements. Groundwater DO was found to increase during the winter and spring months when the site hydraulic gradient increased due to oxygenated recharge from atmospheric precipitation during this period.
Thus, a second biological removal mechanism of chlorinated organic com- pounds may have been occurring at the site, which involved the oxidation of TCE to ethane via methane oxidizing bacteria. Since both ethane and VC-ethylene are all present as possible degradation products of the TCE, it is probable that both mechanisms were operative.
Skubalet al. (1999)investigated temporal changes in redox zonation at a mixed hydrocarbon/solvent contaminated aquifer in an effort to quantify the propensity for TCE natural attenuation in situ. Predominant TEAPs as measured by dissolved hydrogen, suggested temporal variations in reoxy- genation along the plume transect. It was postulated that the intrusion of oxygen was possibly due to recharge, fluctuations in the water table, and/or microbial activity. Microbial analyses revealed a correlation between bacte- rial phylogeny, TEAP, and groundwater hydrogen concentrations. An increase in the water table and evidence of methanogenesis corresponded to an order of magnitude increase in archaeal 16S rRNA relative to when it was unsaturated (creation of capillary fringe). Spatial and temporal variations in TEAPs and microbial community structure suggest that the potential for TCE dechlorination varies seasonally within the plume, with reductive reactions (formation of DCE and VC) more likely in the shallow saturated zone or the capillary fringe during wet cycles, and aerobic co-metabolism of TCE and its products more likely in the deep aerobic subsurface or vadose zone where it could be supported by organic co-contaminants or methane from methanogenesis. Nevertheless, natural dechlorination processes at the site were limited despite the abundance of electron donor and C sources.
Significant vertical and horizontal variations in TEAP zonation and associated organic contaminant degradation processes have been noted in a variety of settings (Christensen et al., 1994; Norris et al., 1994). At a sewage-effluent plume in Cape Cod, MA, the vertical dispersivity was found to be insignificant relative to the longitudinal dispersivity
54 Philip M. Jardine
(Garabedian et al., 1991; LeBlanc et al., 1991). This circumstance limited vertical mixing and allowed for a sharp gradient in oxygen and other solutes within the subsurface over long time periods (Smith et al., 1991) thus influencing contaminant fate and transport processes in the different zones. McGuireet al. (1999) also observed significant temporal variations in TEAP zonation within a mixed fuel/chlorinated solvent plume. Simul- taneous changes in microbial community structure were noted during the same time period as the TEAP shifts, with methanogen abundance being highest where methanogenesis and sulfate reduction were the predominant TEAP (Haack and Reynolds, 1999; Reynolds and Haack, 1999).
Song et al. (2002) used a time-series stable C isotope technique to monitor the degradation of groundwater TCE and its by-products at the Idaho National Laboratory in western USA following subsurface biostimu- lation using lactate. Large kinetic isotope effects were observed during the dechlorination process which indicated a microbially mediated reaction scenario. The observed changes in the 13C/12C ratio indicated microbial- mediated biodegradation of the VOCs, since a constant isotopic ratio would be more indicative of geochemical and hydrological impacts on VOC concentration changes. Morrill et al. (2005) also observed substantial C isotope enrichment in c-DCE at a DoD contaminated site in San Antonio, TX, and the authors were able to calculate the rate of VOC transformation using the stable C isotope technique. Similarly,Chuet al. (2004)suggested that VOC compound-specific isotopic fractionation could assist in deter- mining whether aerobic or anaerobic degradation of VC and c-DCE occur during in situ reductive dechlorination; however, metabolic versus co- metabolic reactions could not be distinguished since their isotopic fraction- ation changes were too similar. Lee et al. (2007) studied stable C isotope fractionation of chloroethenes by dehalorespiring isolates and found that a wide range of isotopic enrichment factors were associated with function- ally similar and phylogenetically diverse organisms. Because of this, the authors cautioned that although compound-specific isotope fractionation is a powerful tool for evaluating the progress ofin situbioremediation in the field, care must be exercised when applying enrichment factors for the interpretation of dechlorination results.
Numerous investigations have shown that fluctuations in recharge to shallow contaminant plumes can create temporal variability in geochemical conditions that are reflected in microbial population changes. Recharge events that deliver electron acceptors such as O2, NO3, SO4, and Fe(III) to anaerobic, contaminated subsurface environments are likely important con- siderations for assessing the propensity for organic contaminant natural attenuation (Vroblesky and Chapelle, 1994). McGuireet al. (2005) noted recharge-induced geochemical changes in a sandy aquifer contaminated with waste fuels and chlorinated solvents. Multiple regression analysis indicated that dominant chemical associations and their interpreted
Influence of Coupled Processes on Contaminant Fate 55