Geochemical reactions resulting from in situ oxidation of PCE-DNAPL by KMnO4 in a sandy aquifer

Although the potential for KMnO4 to destroy chlorinated ethenes in situ was first recognized more than a decade ago, the geochemical processes that accompany the oxidation have not previously been examined. In this study, aqueous KMnO4 solutions (10-30 g/L) were injected into an unconfined sand aquifer contaminated by the dense non-aqueous-phase liquid (DNAPL) tetrachloroethylene (PCE). The effects of the injections were monitored using depth-specific, multilevel groundwater samplers, and continuous cores. Two distinct geochemical zones evolved within several days after injection. In one zone where DNAPL is present, reactions between KMnO4 and dissolved PCE resulted in the release of abundant chloride and hydrogen ions to the water. Calcite and dolomite dissolved, buffering the pH in the range of 5.8-6.5, releasing Ca, Mg, and CO2 to the pore water. In this zone, the aqueous Ca/Cl concentration ratio is close to 5:12, consistent with the following reaction for the oxidation of PCE in a carbonate-rich aquifer: 3C2Cl4 + 5CaCO3(s) + 4KMnO4 + 2H+ --> 11CO2 + 4MnO2(s) + H2O + 12Cl- + 5Ca2+ + 4K+. In addition to Mg from dolomite dissolution, increases in the concentration of Mg as well as Na may result from exchange with K at cation-exchange sites. In the second zone, where lesser amounts of PCE were present, KMnO4 persisted in the aquifer for more than 14 months, and the porewater pH increased graduallyto between 9 and 10 as a resultof reaction between KMnO4 and H2O. A small increase in SO4 concentrations in the zones invaded by KMnO4 suggests that KMnO4 injections caused oxidation of sulfide minerals. There are important benefits of carbonate mineral buffering during DNAPL remediation by in situ oxidation. In a carbonate-buffered system, Mn(VII) is reduced to Mn(IV) and is immobilized in the groundwater by precipitating as insoluble manganese oxide. Energy-dispersive X-ray spectroscopy analyses of the manganese oxide coatings on aquifer mineral grains have detected the impurities Al, Ca, Cl, Cu, Pb, P, K, Si, S, Ti, U, and Zn indicating that, similar to natural systems, precipitation of manganese oxide is accompanied by coprecipitation of other elements. In addition, the consumption of excess KMnO4 by reaction with reduced minerals such as magnetite will be minimized because the rates of these reactions increase with decreasing pH. Aquifer cores collected after the KMnO4 injections exhibit dark brown to black bands of manganese oxide reaction products in sand layers where DNAPL was originally present. Mineralogical investigations indicate that the manganese oxide coatings are uniformly distributed over the mineral grains. Observations of the coatings using transmission electron microscopy indicate that they are on the order of 1 microm thick, and consequently, the decrease in porosity through the formation of the coatings is negligible.     

Groundwater acidification and the mobilization of trace metals in a sandy aquifer

The acidification of groundwater due to acid rain impact and the mobilization of the trace metals Ni, Be, Cd and Co was studied in a noncalcareous sandy aquifer. The groundwater is acidified down to pH 4.4 in the upper 3-4 m of the saturated zone. There is a sharp acidification front and below that the pH increases to 5.2-6.5. The acid zone groundwater contains an Al concentration of approximately 0.2 mM. These observations could be explained by a reactive transport model for downward groundwater movement based on ion exchange and equilibrium with Al(OH)3. At the acidification front, the Al3+ in groundwater exchanges for sorbed Ca2+ and Mg2+ and the coupled dissolution of Al(OH)3 causes the pH to increase. The downward migration rate of the acidification front is 3.5-5.0 cm/yr. Trace metals (Ni, Be, Cd and Co) are found to accumulate near the acidification front. Downward moving, low pH, and trace metal containing groundwater passes the acidification front, and the trace metals adsorb as the pH increases. The acidification front moves downward at a slower rate, and in this process the heavy metals are desorbed. Accordingly, the acidification front functions as a geochemical trap where trace metals accumulate, and their amount will increase with time. Different surface complexation models were explored to explain the behavior of Ni. Neither a simple iron oxide surface complexation model nor ion exchange could explain the field observations of the Ni distribution. The sediment appeared, even at low pH, to have a much stronger affinity toward Ni than predicted by the iron oxide model. The discrepancy can be accounted for in the model by increasing the Ni binding strength constant in combination with an increased number of reactive sites.     

Sorption kinetics of organic contaminants by sandy aquifer and its kerogen isolate

Long-term sorption behaviors of phenanthrene (Phen) on the Borden sand from 1 min to 365 days and of Phen and 1,2-dichlorobenzene (DCB) on the isolated kerogen from 1 to 120 days were characterized by examining the time dependence of solute phase distribution relationships (PDRs) and compared with the prior reported sorption of tetrachlorobenzene (TeCB) and tetrachloroethene (PCE) on the pulverized and/or acid-treated bulk sand and size fractions. The sorption kinetics for Phen on the bulk sand and its kerogen isolate can well be described by the fractional power kinetics equation (q(e) = kt(b)). The similar rate parameter b for q(e)(t) vs t at 5-7 levels of initial concentrations of Phen on the two sorbents, respectively, ranging from 0.077 to 0.099 and from 0.072 to 0.086, indicates the similar sorption kinetics rate. The modified-Freundlich parameters of TeCB on the pulverized or acid-treated 0.3-mm size fraction match those of Phen on the isolated kerogen, suggesting the same natural organic matter (NOM) property of the two sorbents. As the prior investigation underestimated the Koc value for the TeCB sorption on the acid-treated 0.3-mm size fraction by a factor of 1.76, the estimated time to reach 95% of sorption equilibrium is much longer than the prior estimation (over 10 years vs about 2.5 years). The estimated times to reach 95% of sorption equilibrium at three levels of relative solubility for Phen on the bulk sand and its isolated kerogen are, respectively, longer than one decade, demonstrating the similar diffusion length for Phen on the two sorbents. The observed slow sorption kinetics is related to nanometer-pore diffusion within kerogen matrix. The investigation supplies new clues for explaining the often observed much longer persistence of organic contaminants in soils and sediments than the prediction based on the short-term laboratory experiment.     

In situ metal precipitation in a zinc-contaminated, aerobic sandy aquifer by means of biological sulfate reduction

The applicability of in situ metal precipitation (ISMP) based on bacterial sulfate reduction (BSR) with molasses as carbon source was tested for the immobilization of a zinc plume in an aquifer with highly unsuitable initial conditions (high Eh, low pH, low organic matter content, and low sulfate concentrations), using deep wells for substrate injection. Batch experiments revealed an optimal molasses concentration range of 1-5 g/L and demonstrated the necessity of adding a specific growth medium to the groundwater. Without this growth medium, even sulfate, nitrogen, phosphorus, and potassium addition combined with pH optimization could not trigger biological sulfate reduction. In column experiments, precipitation of ZnS(s) was induced biologically as well as chemically (by adding Na2S). In both systems, zinc concentrations dropped from about 30 mg/L to below 0.02 mg/L. After termination of substrate addition the biological system showed continuation of BSR for at least 2 months, suggesting the insensitivity of the sulfate reducing system for short stagnations of nutrient supply, whereas in the chemical system an immediate increase of Zn concentrations was observed. A pilot experiment conducted in situ at the zinc-contaminated site showed a reduction of zinc concentrations from around 40 mg/L to below 0.01 mg/L. Termination of substrate supply did not result in an immediate stagnation of the BSR process, but continuation of BSR was observed for at least 5 weeks.     

Leaching of metribuzin metabolites and the associated contamination of a sandy Danish aquifer

As degradation products of metribuzin have received little attention as potential groundwater contaminants, we evaluated leaching of metribuzin and its primary metabolites desaminometribuzin (DA), desaminodiketometribuzin (DADK), and diketometribuzin (DK) at a sandy test site in Denmark. Soil water and groundwater were sampled monthly over a four-year period. Leaching of metribuzin and DA was negligible. DK and DADK leached from the root zone (1 meter below ground surface (mbgs)) in average concentrations considerably exceeding the EU limit value for drinking water (0.1 microg/L). Both metabolites appear to be relatively stable and persisted in soil water and groundwater several years after application. Past application of metribuzin at the site had contaminated the groundwater with both DK and DADK, which were detected in 99% and 48%, respectively, of the groundwater samples analyzed. Except for three of the groundwater samples, the DADK concentration never exceeded the EU limit value. In contrast, the annual concentration of DK exceeded 0.1 microg/L at 90% of the screens analyzed. The present findings suggest that as the degradation products of metribuzin can leach through sandy soil in high concentrations, they could potentially contaminate the groundwater. In view of this risk DK and DADK should both be included in monitoring programs and their ecotoxicological effects should be further investigated.     

Removal of bacteriophages MS2 and phiX174 during transport in a sandy anoxic aquifer

The objectives of our study were to determine (i) removal of bacteriophage MS2 and phiX174, as surrogates for human pathogenic viruses, in an anoxic aquifer and (ii) the safe length of the microbial protection zone in anoxic aquifers. 3.5 Log units of MS2 were removed by adsorption and inactivation during 63 days residence time, which was 1.4 log units lower than removal of phiX174. These removal rates were considerably lower than previously reported for MS2 and phiX174 in oxic aquifers and consequently longer protection zones around anoxic aquifers might be needed. Therefore, the observed log removal of MS2 was used in a risk assessment approach to determine the required safe length of the microbial protection zone. In case of a leaking sewer in the vicinity of a well in an anoxic aquifer, the risk assessment demonstrated that a microbial protection zone of 110 m may be needed to meet a risk of infection of 10(-4) persons per year. This length can be two to three times larger than the length of the protection zone currently used in a number of countries.     

Carbon and chlorine isotope ratios of chlorinated ethenes migrating through a thick unsaturated zone of a sandy aquifer

Compound-specific isotope analysis (CSIA) can potentially be used to relate vapor phase contamination by volatile organic compounds (VOCs) to their subsurface sources. This field and modeling study investigated how isotope ratios evolve during migration of gaseous chlorinated ethenes across a 18 m thick unsaturated zone of a sandy coastal plain aquifer. At the site, high concentrations of tetrachloroethene (PCE up to 380 μg/L), trichloroethene (TCE up to 31,600 μg/L), and cis-1,2-dichloroethene (cDCE up to 680 μg/L) were detected in groundwater. Chlorinated ethene concentrations were highest at the water table and steadily decreased upward toward the land surface and downward below the water table. Although isotopologues have different diffusion coefficients, constant carbon and chlorine isotope ratios were observed throughout the unsaturated zone, which corresponded to the isotope ratios measured at the water table. In the saturated zone, TCE became increasingly depleted along a concentration gradient, possibly due to isotope fractionation associated with aqueous phase diffusion. These results indicate that carbon and chlorine isotopes can be used to link vapor phase contamination to their source even if extensive migration of the vapors occurs. However, the numerical model revealed that constant isotope ratios are only expected for systems close to steady state.     

Solvent release into a sandy aquifer. 2. Estimation of DNAPL mass based on a multiple-component dissolution model

A chlorinated solvent mixture (2.0 L of trichloroethylene, 0.5 L of chloroform, and 2.5 L of tetrachloroethylene) was released into a sandy aquifer to create a heterogeneously distributed DNAPL (dense nonaqueous-phase liquid) source. The dissolution and dissolved-phase plume development from this source were studied in detail along a cross-section downgradient of the source for a period of approximately 1 year. At the conclusion of the experiment, the site was excavated to map the actual distribution of solvent residuals in the subsurface. Multiple-component dissolution theory provides a tool for the estimation of the mass of a multiple-component DNAPL source present in the groundwater. Concentration ratios between the compounds change with time, and those changes can be used to estimate the mass of DNAPL upgradient of the monitoring point(s) or well(s). The method is independent of the dilution occurring in the groundwater and only requires observations of time series of the contaminants in one or more monitoring points. For the field experiment, the method was applied using the measured concentrations of individual sampling points, the depth-integrated concentrations, the area-integrated concentrations, and the effluent concentrations of the cell. The experiment showed that multiple-component dissolution theory may be a valuable tool for the estimation of the mass of multiple-component DNAPL residuals in the saturated zone.     

Reaction fronts in brick-sand layers: column experiments and modeling

Admixing waste materials with common raw materials in brick production is a promising treatment technology to overcome contamination problems, because organic pollutants are destroyed and inorganic contaminants are thought to be immobilized. During their use in constructions and after the use as part of the demolition masses bricks can be leached by runoff waters and seepage waters. A possible application of recycling crushed bricks consists of their use as a surface layer material on sports grounds or in road construction. To investigate the potential leaching during acidification of a brick-sand layer and the resultant leaching of heavy metals, crushed material from two bricks was examined in several column experiments. Deionized water at pH 4 percolated through the water-saturated columns at a Darcy velocity which was varied between 0.37 and 2.2 m/d. Another column was run under unsaturated conditions. A reaction front evolved in all experiments characterized by a pH increase from pH 4 to pH 8. The chemical composition of the percolating water changed at the reaction front. Several heavy metals (Cd, Co, Cu, Ni) and Al were immobilized at this front. Other parameters such as Ca, S as SO4, V, and Mo were depleted within several days. The reaction front moved forward depending on the Darcy velocity in the column and the buffer capacity of the brick sand. Thermodynamic calculations (PHREEQC 2.0) indicated that mobilization of As was influenced by Ba(AsO4)2. The solubility of Ba and Mn was controlled by barite and manganite, respectively. Reactive transport modeling was applied to describe the dissolution of the bricks with regard to their main components Ca, SO4, Al, and Si.     

Interpreting deposition patterns of microbial particles in laboratory-scale column experiments

The transport and fate of microbial particles in subsurface environments is controlled by their capture (natural filtration) by sediment grains. Typically, filtration models used to describe microbe removal in porous media predict exponential decrease in microbial particle concentration with travel distance. However, a growing body of laboratory-scale column experiments suggests that the retained microbial particle profiles decay nonexponentially. The observed behavior may be attributed to the heterogeneity in the interactions between microbial particles and sediment grains, most likely due to the inherent variability in the microbial particles. This factor can be incorporated into classical colloid filtration (deposition) theory by inclusion of a distribution in the deposition rate coefficient. We show that certain distributions of the deposition rate coefficient (i.e., log-normal, bimodal, and power-law distributions) give rise to nonexponential deposition patterns. Comparisons of model predictions to experimental data indicate that the observed nonexponential deposition behavior of bacteria and virus particles may be attributed to a broad range (i.e., a power-law distribution) of microbial deposition rates. Other mechanisms such as particle release and blocking by previously deposited microbial particles are also shown to be potential sources of deviation from the classical filtration theory. Our results further suggest that monitoring fluid-phase particle concentration is insufficient for accurate characterization of the deposition and transport behavior of microbial particles in saturated porous media. Rather, the shape of the microbial particle retention profile is shown to be a key indicator of the mechanisms controlling microbial deposition and transport.     

Spatial and temporal observations of adsorption and remobilization of heavy metal ions in a sandy aquifer matrix using Magnetic Resonance Imaging

The exploration of the transport and matrix interactions of heavy metal ions in the subsurface environment under natural conditions is an important field of research in environmental science and technology. Most commonly, column tests are used for a first assessment of the transport behavior. Classical column tests fall short with regard to the spatial and temporal resolution; however, these detailed data are needed for proper upscaling. Hence, providing spatially and temporally resolved data on the distribution of environmentally relevant concentrations of heavy metal ions in a water-wet aquifer matrix poses a major challenge to analytical chemistry. In this contribution, we present the results of Magnetic Resonance Imaging (MRI) studies in which submilligram quantities of heavy metal ions where either fed conventially through the column or locally injected into saturated sand packings. The subsequent transport and mobilization was monitored at a high spatial and temporal resolution. The results from a local injection show that the test design of column tests has not yet come to an end and that column tests under MRI-control may be used as a model system for, e.g., remediation techniques.     

Transport and bacterial interactions of three bacterial strains in saturated column experiments

The impact of bacteria-solid and bacteria-bacteria interactions on the transport of Klebsiella oxytoca, Burkholderia cepacia G4PR1, and Pseudomonas sp. #5 was investigated in saturated sand column experiments (L = 114 mm; ? = 33 mm) under constant water velocities (～ 5 cm · h(-1)). Bacterial strains were injected into the columns as pulses either individually, simultaneously, or successively. A one-dimensional mathematical model for advective-dispersive transport and for irreversible and reversible bacterial kinetic sorption was used to analyze the bacterial breakthrough curves. Different sorption parameters were obtained for each strain in each of the three experimental setups. In the presence of other bacteria, sorption parameters for B. cepacia G4PR1 remained similar to results from individual experiments, indicating the presence of other bacteria generally had a lesser influence on its migration than for the other bacteria. K. oxytoca is more competitive for the sorption sites when simultaneously injected with the other bacteria. Ps. sp. #5 generally yielded the greatest detachment rates and the least affinity to attach to the sand, indicative of its mobility in groundwater systems. The results of this study clearly indicate both bacteria-solid and bacteria-bacteria interactions influence the migration of bacteria. A more complete understanding of such interactions is necessary to determine potential migration in groundwater systems.     

Passive in situ remediation of metal-polluted water with caustic magnesia: evidence from column experiments

Passive remediation consists of a permeable system that enables the water to pass through while retaining metals by means of biogeochemical reactions. Conventional passive treatments are based on calcite dissolution. This increases the pH to values between 6 and 7, which are insufficiently high to precipitate divalent metals. Alternative treatments are based on sulfate reduction with organic matter in order to precipitate metal sulfides. However, redox reactions are usually too slow to treat large groundwater flows as currently found in gravel aquifers (>50 m/a). Caustic magnesia obtained from calcination of magnesium carbonate was tested as an alternative material to devising passive remediation systems. Caustic magnesia reacts with water to form magnesium hydroxide, which dissolves, increasing the pH to values higher than 8.5. Then zinc and lead are mainly precipitated as hydroxides, copper is precipitated as hydroxysulfate, and manganese(II) is oxidized and precipitated as manganese(III) oxides. Thus, metal concentrations as high as 75 mg/L in the inflowing water are depleted to values below 0.04 mg/L. Magnesia dissolution is sufficiently fast to treat flows as high as 100 m/a. The new precipitates may lead to a permeability drop in the porous treating system. Mixtures of caustic magnesia and an inert material such as silica sand (approximately 50% of each) have been shown to be as reactive as pure magnesia and permeable for a longer time (more than 10 months and 1000 pore vol).     

Differential effects of dissolved organic carbon upon re-entrainment and surface properties of groundwater bacteria and bacteria-sized microspheres during transport through a contaminated, sandy aquifer

Injection-and-recovery studies involving a contaminated, sandy aquifer (Cape Cod, Massachusetts) were conducted to assess the relative susceptibility for in situ re-entrainment of attached groundwater bacteria (Pseudomonas stuzeri ML2, and uncultured, native bacteria) and carboxylate-modified microspheres (0.2 and 1.0 μm diameters). Different patterns of re-entrainment were evident for the two colloids in response to subsequent injections of groundwater (hydrodynamic perturbation), deionized water (ionic strength alteration), 77 μM linear alkylbenzene sulfonates (LAS, anionic surfactant), and 76 μM Tween 80 (polyoxyethylene sorbitan monooleate, a very hydrophobic nonionic surfactant). An injection of deionized water was more effective in causing detachment of micrsopheres than were either of the surfactants, consistent with the more electrostatic nature of microsphere's attachment, their extreme hydrophilicity (hydrophilicity index, HI, of 0.99), and negative charge (zeta potentials, ζ, of -44 to -49 mv). In contrast, Tween 80 was considerably more effective in re-entraining the more-hydrophobic native bacteria. Both the hydrophilicities and zeta potentials of the native bacteria were highly sensitive to and linearly correlated with levels of groundwater dissolved organic carbon (DOC), which varied modestly from 0.6 to 1.3 mg L(-1). The most hydrophilic (0.52 HI) and negatively charged (ζ -38.1 mv) indigenous bacteria were associated with the lowest DOC. FTIR spectra indicated the latter community had the highest average density of surface carboxyl groups. In contrast, differences in groundwater (DOC) had no measurable effect on hydrophilicity of the bacteria-sized microspheres and only a minor effect on their ζ. These findings suggest that microspheres may not be very good surrogates for bacteria in field-scale transport studies and that adaptive (biological) changes in bacterial surface characteristics may need to be considered where there is longer-term exposure to contaminant DOC.     

Manganese oxidation induced by water table fluctuations in a sand column

On-off cycles of production wells, especially in bank filtration settings, cause oscillations in the local water table, which can deliver significant amounts of dissolved oxygen (DO) to the shallow groundwater. The potential for DO introduced in this manner to oxidize manganese(II) (Mn(II)), mediated by the obligate aerobe Pseudomonas putida GB-1, was tested in a column of quartz sand fed with anoxic influent solution and subject to 1.3 m water table changes every 30-50 h. After a period of filter ripening, 100 μM Mn was rapidly removed during periods of low water table and high dissolved oxygen concentrations. The accumulation of Mn in the column was confirmed by XRF analysis of the sand at the conclusion of the study, and both measured net oxidation rates and XAS analysis suggest microbial oxidation as the dominant process. The addition of Zn, which inhibited GB-1 Mn oxidation but not its growth, interrupted the Mn removal process, but Mn oxidation recovered within one water table fluctuation. Thus transient DO conditions could support microbially mediated Mn oxidation, and this process could be more relevant in shallow groundwater than previously thought.