Grantee Research Project Results
Final Report: Bioavailability of Aromatic Hydrocarbons in Saturated Porous Media: The Effects of Chemical Aging and Mass Transfer
EPA Grant Number: R825406Title: Bioavailability of Aromatic Hydrocarbons in Saturated Porous Media: The Effects of Chemical Aging and Mass Transfer
Investigators: Bouwer, Edward J. , Ball, William P.
Institution: The Johns Hopkins University
EPA Project Officer: Chung, Serena
Project Period: December 6, 1996 through December 5, 1999 (Extended to December 5, 2001)
Project Amount: $439,725
RFA: Environmental Fate and Treatment of Toxics and Hazardous Wastes (1996) RFA Text | Recipients Lists
Research Category: Hazardous Waste/Remediation , Land and Waste Management , Safer Chemicals
Objective:
This research effort was designed to explore the hypothesis that effects of aging on contaminant desorption and biodegradation can be directly related to processes of contaminant diffusion from regions of the sorbent that are inaccessible to (or at least unoccupied by) the microorganisms mediating the degradation. In particular, the driving hypotheses of this research were that: (1) mass transfer effects are rate controlling under many natural circumstances; (2) aging can reduce rates of release to bacterially accessible regions; and (3) the rates of biodegradation of aged materials can be mechanistically quantified by understanding the intrinsic (bulk-aqueous-phase) rate of biodegradation, the equilibrium conditions that drive contaminant mass distribution (sorption isotherm characteristics), and the parameters that control rates of intrasorbent diffusion. A related hypothesis was that these various parameters could be incorporated into an overall modeling approach that will allow accurate simulation of model reactor systems and provide more accurate prediction of the effect of contaminant/solid exposure time ("aging" period) on intrinsic biodegradation rates in natural aqueous environments.
In consideration of the above hypotheses, the objectives of this research were to:
- Evaluate, quantify, and model the combined effects of sediment/water partitioning (sorption) and intrasorbent diffusion on the bioavailability and biodegradation of selected hydrophobic organic contaminants (HOCs) under conditions simulating a heterogeneous aquifer environment where sediments have been exposed to long-term contamination; and
- Quantify the contaminant "bioavailability" in terms of fundamental properties of the contaminant/water/solid systems, the intrinsic rates of bioremediation in aqueous solution, and the contact period for aged systems.
Summary/Accomplishments (Outputs/Outcomes):
This research included abiotic and biotic experimental studies as well as associated modeling efforts and data interpretation. Experimental efforts have included batch biodegradation, sorption, and desorption studies as well as column-based transport experiments through packed beds of the various sediments. In both the batch and column work, emphasis has been on comparison of desorption or biodegradation rates under conditions controlled by desorption, with particular emphasis on comparison of rates after differing periods of "aging" (differing periods of sediment exposure to aqueous-phase contaminants). Natural geosorbents studied included both coarse-grained and fine-grained materials from three different locations. These included a previously studied sandy soil from Bozeman, MT; a well-characterized aquifer sand from Borden, Ontario (having intraparticle porosity and very low carbon content); and a previously studied fine-grained clayey silt material from a groundwater aquitard at Dover Air Force Base, DE (hereafter referred to as orange silty clay loam, or OSCL). In addition, control experiments were conducted with combusted Ottawa sand (quartz sand). Descriptions of these experiments along with a summary of the important findings are provided below.
Biotic Batch Mineralization Studies
Batch mineralization studies were conducted in liquid cultures and with aged sediments using naphthalene and phenanthrene as substrates for a previously enriched microbial population obtained from a coal tar contamination site. Batch aging/mineralization studies in the presence of Ottawa sand, Borden sand, Dover OSCL solids, and Bozeman sediments were undertaken to evaluate the effect of sorption and contaminant exposure time on the initial mineralization rates of naphthalene and phenanthrene. Aging periods investigated included 1, 7, 30, 90, 210, 270, and 900 days.
These studies revealed that the initial naphthalene and phenanthrene mineralization rates in liquid cultures were 2 to 30 times greater than those observed in the sediment slurry systems when comparing systems with equivalent initial contaminant mass. The rate and extent of mineralization were a function of aging time with Bozeman sediments and Borden sand, but not with Dover OSCL. These data illustrate that the presence of solids and contaminant exposure time can impact the rate of biodegradation of HOCs such as naphthalene and phenanthrene in the presence of natural geosorbents under saturated conditions. The data also provide a well-controlled set of results for the development and evaluation of numerical models to simulate the combined effects of sorption, mass transfer, and biodegradation.
Batch Equilibrium and Rate Studies
A series of batch equilibrium and rate studies were conducted to evaluate the rate and extent of sorption and desorption of naphthalene and phenanthrene in the presence of the three natural geosorbents. Batch studies were conducted to develop equilibrium isotherms for all sorbate/sorbent combinations and to obtain sorption/desorption rates under different aging conditions. The primary objectives of these studies were to: (1) evaluate the impact of abiotic mixing time on attainment of sorption equilibrium for naphthalene and phenanthrene in the presence of natural geosorbents; (2) evaluate the impact of abiotic mixing times, or aging, on the rates of naphthalene and phenanthrene desorption from natural geosorbents; and (3) apply equilibrium and rate models to the batch study results.
Abiotic Batch Equilibrium Studies. Batch sorption equilibrium experiments were completed with naphthalene and phenanthrene using Borden sand, Ottawa sand, Dover OSCL, and Bozeman sediments. Multiple experiments were conducted to determine the time necessary to reach true equilibrium with each of the sorbent/sorbate combinations. Mixing periods evaluated were 7, 30, 90, and 270 days.
Both linear and nonlinear (Freundlich) sorption equilibrium models were used to evaluate the results. A summary of these results is provided in Table 1. As demonstrated in Table 1, sorption equilibrium was achieved rapidly for Dover OSCL, with true equilibrium achieved in less than
7 days. Continued uptake of naphthalene and phenanthrene by Borden sand and Bozeman sediment was observed throughout the 270-day mixing period. Considering the full range of concentration data, all results could be fit by Freundlich isotherm relations, with values of the Freundlich coefficient (1/n) ranging from 0.75 to 0.96. If only partial data sets at higher concentrations are considered, greater isotherm linearity was observed.
To aid in data interpretation and modeling efforts, batch sorption experiments were conducted using pulverized Borden sand and Bozeman sediments to determine the contaminant distribution under conditions that are expected to closely approximate equilibrium. (Pulverization shortens the diffusive length scale and reduces the mixing time needed to reach true equilibrium.) The average distribution coefficient (Kd) was 13.1 ± 2.0 L/Kg for Borden sand with phenanthrene and 145 ± 19 L/Kg for Bozeman material and phenanthrene, which is in good agreement with results for unpulverized materials aged for 270 days (Table 1). Therefore, equilibrium conditions had likely been achieved by the 270-day mixing period.
Sorbent | Average Kd Values (L/kg) for Geosorbent Naphthalene | |||
---|---|---|---|---|
7-Day | 30 Day | 90-Day | 270-Day | |
Borden | 0.35 ± 0.11 | 0.42 ± 0.02 | 0.42 ± 0.02 | 0.52 ± 0.03 |
Dover OSCL | 0.73a | 0.68 ± 0.08 | 0.62 ± 0.2 | 0.72 ± 0.12 |
Bozeman | 4.0 ± 0.9 | 4.9 ± 0.7 | 5.2 ± 0.9 | 6.3 ± 1.2 |
Ottawa | NSb | NSb | NSb | NSb |
Sorbent | Average Kd Values (L/kg) for Geosorbent and Phenanthrene | |||
7-day | 30-Day | 90-Day | 270-Day | |
Borden | 4.31 ± 0.9 | 5.04 ± 0.8 | 11.1 ± 2.6 | 13.4 ± 2.1 |
Dover OSCL | 14.5a | 14.1 ± 1.0 | 13.6 ± 0.9 | 15.6 ± 1.6 |
Bozeman | 9 ± 2.0 | 23 ± 4.4 | 135 ± 26 | 155 ± 31 |
Ottawa | NSb | NSb | NSb | NSb |
a Data from Xia (1998)
b NS = No measurable sorption; observed Kd values were less than 0.01. Aqueous concentration ranges for phenanthrene were 0.015 mg/L to 0.9 mg/L, while the aqueous concentration range for naphthalene was 0.024 mg/L to 27 mg/L.
Abiotic Batch Desorption Rate Studies. Abiotic desorption rate experiments were conducted using naphthalene and phenanthrene with the Borden sand, Dover OSCL, and Bozeman sediments. The objective was to quantify the rate of naphthalene or phenanthrene desorption from natural geosorbents that had previously been exposed to these sorbates for different aging periods. The aging periods investigated were 7, 30, 90, 270, and 900 days. Desorption was evaluated by means of an "infinite-sink" desorption technique modified from Pignatello (1990) and Cornelissen et al. (1997). The method relies upon the presence of a strong synthetic sorbent (TenaxO) to maintain negligible chemical activity in the aqueous phase. The mass of contaminants desorbed from the soils and sediments accumulates in the Tenax phase, where it is subsequently quantified by means of extraction into a hexane phase containing a known concentration of bromobenzene as internal standard.
The desorption response observed in these studies followed a typical pattern of rapid desorption followed by much slower desorption over a prolonged period. The rate and extent of the more rapidly desorbing portion was a function of aging time for the Borden sand and the Bozeman sediments but not for the Dover OSCL sediments. For the Borden and Bozeman material, fractional rates of removal ((dS/dt)/So) were roughly independent of aging time after about 200 hours (Bozeman). However, the fraction of initially sorbed material that was subject to this slower rate was higher for the more extensively aged samples. These findings are consistent with the concept that a higher fraction of material had diffused more deeply into the sorbent with the longer aging times. Effects were more pronounced with phenanthrene than with naphthalene, which is consistent with our research hypothesis that the diffusion process is influenced by intrasorbent retardation.
The desorption data were interpreted using the following two rate models:
First Order Model:
(1)
where:
S = sorbed phase concentration [mg/Kg]
km = first-order mass transfer coefficient [hr-1]
Two-Compartment Rate Model:
(2)
where:
St = sorbed phase concentration at time, t [mg/Kg]
So = initial sorbed phase concentration [mg/Kg]
Frap = rapidly desorbing fraction
Fslow = slowly desorbing fraction
km,rap = rapid desorption rate constant [hr-1]
km,slow = slow desorption rate constant [hr-1]
The two-compartment model provided a superior fit to the data. This trend was consistent for all of the desorption results. Results for the fitted two-compartment model for naphthalene and phenanthrene experiments are summarized in Tables 2 and 3. The square of the correlation coefficient (R2) for each regression analysis also is given in these tables. It is evident from the data presented in these tables that desorption is more rapid from the Dover OSCL sediments than the other geosorbents. The effect of aging on parameter estimation is illustrated by the observed reduction in the parameter km,slow when the aging time was increased from 7 to 270 days. Trends in these data are consistent with our expectations based on concepts of diffusion limited desorption and an initial (nonequilibrium) distribution of contaminants (which changes form as a function of aging time). Alternative (more mechanistic) modeling will therefore require proper consideration of both sorption nonlinearity and diffusion rates. These data provide a sound foundation for such modeling, which is being explored as part of an ongoing effort.
Geosorbent | Naphthalene | Phenanthrene | ||||||
---|---|---|---|---|---|---|---|---|
R2 | Fslow | km,rap (hr-1) | Km, slow (hr-1) | R2 | Fslow | Km,rap (hr-1) | Km,slow (hr-1) | |
Dover OSCL | 0.994 | 0.215 | 12.37 | 0.0083 | 0.988 | 0.190 | 6.18 | 0.0062 |
Borden Sand | 0.997 | 0.412 | 0.233 | 0.0034 | 0.996 | 0.560 | 0.309 | 0.0038 |
Bozeman Sediments | 0.979 | 0.224 | 1.85 | 0.049 | 0.989 | 0.380 | 0.813 | 0.0020 |
Geosorbent | Naphthalene | Phenanthrene | ||||||
---|---|---|---|---|---|---|---|---|
R2 | Fslow | km,rap (hr-1) | Km, slow (hr-1) | R2 | Fslow | Km,rap (hr-1) | Km,slow (hr-1) | |
Dover OSCL | 0.977 | 0.237 | 50 | 0.0020 | 0.978 | 0.208 | 3.26 | 0.0018 |
Borden Sand | 0.998 | 0.441 | 0.119 | 0.0011 | 0.998 | 0.621 | 0.045 | 0.0006 |
Bozeman Sediments | 0.992 | 0.396 | 4.77 | 0.0007 | 0.995 | 0.464 | 0.457 | 0.0008 |
Summary of Batch Study Results. Sorption studies revealed that the apparent Kd increased with contact time for Borden sand and Bozeman sediments, but changes in Kd were not observed for Dover OSCL solids. Desorption rate studies also were consistent with these findings?for the Bozeman and Borden sediments, increasing times of sorption exposure lead to substantial decreases in the amount of desorption that can be achieved within the first 1,000 hours. Thus, the reduced mineralization in the two sediments with aging time can be directly related to issues of abiotic mass transfer.
These results confirm that sorption/desorption processes and contaminant aging can significantly impact the bioavailability of HOCs in contaminated sediments and that these effects can be directly related to mass transfer limitations. In particular, the results are consistent with the hypothesis that failure to attain equilibrium during a forward diffusive process (sorption) can lead to distributions of contaminant under which a larger fraction of mass is initially accessible for rapid release. Important implications of the model, however, are that no contaminant is "permanently" sequestered, and that the ultimate rates of final desorptive release are controlled by the same physical processes for both aged and unaged samples.
Bioavailability Model
Prior work in our laboratories has shown how the effects of both sorption equilibrium and rate can be accounted for through the application of a "bioavailability factor" approach. The modeling approach shown below is based on simplifying assumptions of linear partitioning, first-order biodegradation, and first-order desorptive mass transfer.
where rb = rate of biological transformation [mg/L-day]; Bf = bioavailability factor [-]; kb = first order biodegradation rate constant [day-1]; and C = aqueous contaminant concentration [mg/L].
The Bioavailability Factor is defined as follows:
where:
Here, = Thiele Modulus [-]; km = first order mass transfer rate coefficient [day-1]; Kd = linear equilibrium partition coefficient [L/kg]; and Rs/w = soil-water ratio [kg/L]. can be thought of as a ratio of the characteristic times for desorption and biodegradation. Bf is thus controlled by the extent of sorption (Kd, Rs/w) as well as by relative rates of mass transfer and biodegradation (). At small Bf, sorption significantly limits biodegradation. The relative importance of mass transfer rates on Bf is dependent on the magnitude of . Large values of (> 1) indicate that mass transfer limits the bioavailability of a compound. All of the parameters comprising Bf and can be determined independently and together provide a useful framework in which to evaluate the impact of sorption on biodegradation.
Parameters developed from the abiotic and biotic batch studies (Kd, km, and kb) were incorporated into the Bf model to predict the effect of natural geosorbents (Dover OSCL, Borden sand, and Bozeman soil) and contaminant aging on the removal of phenanthrene. The simulations revealed that the presence of solids can result in up to a seven-fold increase in the time required for 99 percent removal of phenanthrene. In addition, simulations using parameters developed from aging studies (Kd and km) resulted in an additional four- to five-fold increase in the time needed to achieve 99 percent removal of phenanthrene.
Column Studies
Laboratory sediment columns were constructed and long-term experiments are being conducted to evaluate effects of sorption, mass transfer, and contaminant aging on the bioavailability of phenanthrene. These experiments are conducted using precision syringe pumps that maintain steady low flow rates under pulseless conditions, and are conducted in a fully headspace-free setting. These conditions, combined with automated sample collection of radio-labeled solutes, offer a means of obtaining very precise measurement of eluted mass over time. They thereby offer the promise of a more precise estimation of slow long-term desorption rates than can be achieved in batch desorption experiments, which involve transfer of samples among vessels and quantification of desorbed mass on the basis of a two-step transfer and extraction process.
Moreover, the column measurements are made under continuous-flow transport conditions that are directly relevant to field-scale conditions of groundwater movement. In fact, for the case of the silty clay aquitard material (OSCL), the diffusion limitations to sorption and desorption are believed to occur at the interparticle scale (i.e., in compacted and impermeable zones that cannot be maintained in batch systems). In this regard, the column macropore configuration is the only practical means of studying diffusion into a well-characterized and stable matrix of this material (Young and Ball, 1998).
Column experiments have included: (1) tracer studies with tritiated water; (2) breakthrough studies with 14C-labeled phenanthrene, followed by longer term "aging"; and (3) abiotic elution of the 14C-labeled contaminant under clean-water feed conditions. Fourteen columns of three different materials (Borden sand, OSCL/Ottawa macropore, and Ottawa sand) were prepared, saturated with synthetic groundwater, and subjected to study. All studies involve the use of pulse-free high pressure syringe pumps and robotic sample collection in systems comprised solely of stainless-steel and fused silica. Long-term exposure to constant concentrations of 14C-labeled phenanthrene is achieved by means of a presaturation column maintained at 5?C to assure a constant concentration well below the 20?C solubility.
Hydrodynamic Evaluation: Tracer studies were performed using dirac (pulse) inputs of tritiated water to estimate the mean hydraulic residence time and hydraulic residence time distribution. For the sand columns, the observed tracer effluent histories were used to determine the aqueous pore volume (for the known flow rate) and the hydromechanical dispersion coefficient. For these columns, experimentally generated breakthrough curves were compared to simulations produced using an analytical solution (Kreft and Zuber, 1978). For the OSCL/macropore columns, observed tracer effluent histories were used, together with known information about the porosity and dispersivity of the Ottawa sand, to characterize diffusion characteristics in the annular silty clay region. For these columns, experimentally generated breakthrough curves were compared to simulations produced using a numerical transport code that couples advection/dispersion in the macropore with diffusion in the annular region (Young and Ball, 1998). Results for sand columns are as shown in Table 4. In general, the sand column results showed excellent data fit, indicating that nonreactive transport could be well-described using a simple convection/dispersion approach, and using values for hydrodynamic dispersion coefficients that were reasonably constant among replicate samples (within a factor of two) and consistent with expectations based on prior work (Young and Ball, 1994; Young and Ball, 1997a). Interpretation of the results for OSCL/macropore columns are pending a final destruction of these columns for accurate estimation of macropore porosity. In general, these results also are consistent with prior work (Young and Ball, 1998) and are indicative of advection/dispersion within the sand macropore, coupled with interparticle diffusion as the sole mechanism of transport into the annular OSCL domain.
Borden 1 | Borden 2 | Borden 3 | Borden 4 | Borden 5 | Borden 6 | Ottawa 1 | Ottawa 2 | |
Total PV, mL | 13.23 | 10.3 | 10.6 | 11.1 | 10.4 | 10.55 | 10 | 7.64 |
Experi-mental Run Time, PV | 2.19 | 2.19 | 2.15 | 2.03 | 2.03 | 2.31 | 2.33 | 1.88 |
C/Co | 0.0021 | 0.0048 | 0.0053 | 0.0127 | 0.020 | 0.0124 | 0.0019 | 0.0169 |
Mass Balance, % | 97.2 | 100.1 | 100.0 | 96.7 | 99.7 | 96.8 | 101.5 | 99.9 |
Peclet Number | 155 | 325 | 205 | 200 | 300 | 230 | 185 | 95 |
Column Breakthrough and Elution. Each of the experimental columns was subjected to continuous exposure to a constant concentration of 14C-labeled phenanthrene for a pre-determined aging period, followed by a period of extended elution with contaminant-free synthetic groundwater. A step increase in concentration was provided, and the initial "breakthrough" of this step change was monitored by means of frequent (automatic) collection of effluent samples. After breakthrough, each of the columns was aged for various periods of time ranging from 90 to greater than 350 days and then subjected to elution with clean synthetic groundwater. Characterization of the extent of sorption during the initial breakthrough period was conducted through both moment analysis and numerical transport codes that couple advection/dispersion with diffusion into "immobile" water and sorption regions, such as in the porous sand particles and annular OSCL region. The column design and operation was based on previous successful research with tetrachloroethene (Young and Ball, 1994; Young and Ball, 1998) and 1,2,4-trichlorobenzene (Young and Ball, 1998; Young and Ball, 1999). Operation of the system with phenanthrene was complicated by some experimental difficulties that were uncovered only in the last year of the project, during the elution of the first column. These difficulties include: (1) phenanthrene sorption to stainless-steel tubing that had been subjected to long periods of use with the column apparatus; and (2) erratic phenanthrene sampling when the robotic fraction collector was used in a "dipping" mode. These issues were resolved prior to most elution studies and had no adverse impact on the aging of the columns. Resolution of these problems, however, further delayed the conduct and interpretation of much of the intended column work. The experimental breakthrough and elution data from most columns are now part of ongoing efforts with other sources of funding and, for the most part, are not covered in this final report.
To our knowledge, the experimental difficulties we encountered in our column work with phenanthrene have not been reported in the literature. The following sections, therefore, elaborate on these problems, so that other researchers might avoid similar experimental artifacts in the future.
Studies of Sorption to Small-Bore Stainless Steel Tubing. Our first elution results caused us to suspect that the extent of retardation and spreading of the phenanthrene front during breakthrough and elution was being influenced not only by the porous medium, but also by phenanthrene sorption within the small-diameter stainless-steel tubing that was used at the influent and effluent ends of the columns. A series of breakthrough and elution experiments on this tubing revealed that roughly 0.13 to 0.27 mg of phenanthrene were retained per cm2 of tubing surface. These masses correspond to retardation factors (within the tubing) of between 7 and 14, which would be equivalent to roughly 9 to 18 monolayers of phenanthrene on the tubing wall, if molecularly smooth tubing were to be assumed. Overall, the results suggested the presence of hydrophobic surfaces on the inner walls of the stainless steel. Subsequent solvent cleaning appears to have removed most of the problem and suggests that phenanthrene may have been sorbing to an organic film within the stainless-steel tubing. New stainless-steel tubing did not exhibit similar problematic behavior.
Two types of commercially available fused silica tubing (0.033 cm ID, Alltech, Deerfield, IL) also were investigated with regard to retention of dissolved phenanthrene during aqueous transport. These tubing samples were advertised as having "polar" or "intermediate polar" deactivations of fused silica surfaces. Both were found to exhibit minimal phenanthrene sorption and retention, with the "polar" tubing being the least sorbing. Ongoing and future column studies with phenanthrene will employ the polar fused silica tubing, which will be frequently cleaned with organic solvent (i.e., rinsed with methylene chloride, followed by acetone and water).
Sample Collection Studies. In addition to the tubing effects noted above, effluent concentrations in the preliminary breakthrough curves also reflected both unexpectedly high variability and some unexplained jumps in concentration that further hindered data interpretation. As background, we note that the column effluent is automatically collected into a series of sequential scintillation vials (for subsequent 14C quantification) by means of a robotic sample collector. With our initially applied method, the fraction collector was programmed to automatically "dip" the sample tubing below the surface of scintillation fluid in foil-covered vials to minimize sample loss by volatilization. Long-term samples were typically sampled only periodically (i.e., there typically were some programmed "lag times" between collected samples, with less than 100 percent of the effluent volume collected). (During "lag" periods, the tubing hangs idle over a waste receptacle.)
The aforementioned method has proven to be extremely reliable and precise in prior transport studies using comparatively volatile organic chemicals such as tetrachloroethene (PCE) and trichlorobenzene. However, the aforementioned observations of variability and erratic changes suggested that phenanthrene (which is less volatile and more hydrophobic) may be subject to additional methodological artifacts. Additional studies therefore were undertaken to identify and rectify the problem. Systematic variation of the time between sequential samples (under conditions of steady feed concentration) revealed a correlation between the "lag time" between samples and the variability of observed concentrations in sequential vials. High lag time also was associated with a bias toward increasingly high apparent concentration. The proposed explanation for both effects is that water tends to spread up the external surface of the stainless- steel tubing during the "lag" period. Phenanthrene accumulates on the external tubing surface as this water evaporates, and the phenomenon may be compounded by the presence of residual scintillation cocktail on the tube. (Surfactant in this fluid may both enhance the upward spreading of the water and present a favorable environment for phenanthrene partitioning.) In any case, the accumulated phenanthrene is "rinsed" when the tube is dipped into the subsequent sample. A series of tests showed that the problem is eliminated when the automatic collector mechanism is modified to prevent such "dipping" of the sampling tube (i.e., if this feature of the robotic mechanism is disabled). (Phenanthrene, as opposed to PCE, is not subject to substantial volatilization during drop fall.)
None of the experimental artifacts addressed here have previously been reported in the literature. For this reason, their nature and resolution should be of interest to the research community. Conference presentations of these effects have been made, and publication is being considered.
Modeling Breakthrough and Elution
Numerical models that couple solute transport with diffusion-limited sorption have previously been developed for spherical diffusion and for annular impermeable domains surrounding macropore flow. The development, parameterization, and application of these models have been described in previous publications (Young and Ball, 1995; Young and Ball, 1997; Young and Ball, 1998; Young and Ball, 1999). In this work, these models have been used to provide numerical simulation of solute breakthrough and elution in columns of the Ottawa and Borden sands, as well as in the OSCL/macropore column. More specifically, the models were employed in three ways: (1) during the design stage of the project to determine which experimental conditions would give the most insight into questions of bioavailability; (2) toward the simulation of our breakthrough and elution results and the calibration of modeling parameters; and (3) in a predictive sense, with hypothetical scenarios of transport and biodegradation. As noted above, our column studies have not yet been completed or subjected to final modeling interpretation. Model application has nonetheless served to provide some important insights, and also was instrumental in helping us to identify the aforementioned experimental difficulties. Moreover, the tenets of the model are well supported by the extensive body of batch system bioavailability results obtained from this project and also are consistent with some of our prior abiotic work with these same systems.
In the above context, predictive applications of our proposed coupled model of sorption and diffusion give insight into issues of bioavailability as relevant to both the aquifer (Borden sand) and aquitard (OSCL) systems. For example, model predictions suggest that, if a macropore column is loaded with phenanthrene at 370 mg/L for 1 month and then subjected to "clean-water elution," it will exhibit two-phase desorption behavior that reflects the combination of slow diffusion processes and initial (nonequilibrium) intrasorbent concentration gradients. Model predictions show that, upon flushing with clean water, such a column will exhibit a rapid initial release of that fraction of contaminants that were near the sorbent surface, followed by a very slow "bleed" of the remaining sorbed material, some of which continues to diffuse deeper into the sorption zone. In fact, of the 76.5 mg of phenanthrene initially sorbed in such a column, only 42.3 mg (55 percent) was eluted within the first month of flushing. Of this, 95 percent of the mass was eluted within the first 3 days and, after a month of elution, desorptive flux was at a rate that could only maintain an effluent aqueous concentration of 2.2 mg/L. Similar results are obtained as a result of intragranular diffusion in the Borden sand.
The above results may be typical of any field scale scenario that is comparable to our column systems (i.e., to situations with similar ratios of mobile zone velocities and effective diffusion rate constants, as determined by the dimensions and properties of the sorbing zone domain). Because such scenarios commonly are encountered in the environment, the findings of this work should have direct relevance to bioavailability issues in the field. For example, the results noted above would imply long-term low rates of mass transport into the mobile (bioavailable) ground water region, with aqueous concentrations that might well be too low to satisfy minimum needs for bacterial growth. On the other hand, it also is possible that such mass transport rates may be too low to cause "significant" health concern to downgradient receptors. Overall, modeling results of this type clearly illustrate that issues of "sequestration" and bioavailability can be very case dependent, with "sequestered" material not often as irreversibly bound as some models would predict.
1This scenario was evaluated through a model run using input parameters equivalent to our column conditions for the OSCL/ macropore system: i.e., macropore zone velocity = 545 cm/day, Kd [OSCL/phenanthrene] = 0.027 m3/kg ; macropore Peclet Number = 100, effective pore diffusion coefficient in the immobile region, Dp = 0.1547 x 10 -10 m2/s.
Journal Articles on this Report : 3 Displayed | Download in RIS Format
Other project views: | All 11 publications | 3 publications in selected types | All 3 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Xia G, Ball WP. Adsorption-partitioning uptake of nine low-polarity organic chemicals on a natural sorbent. Environmental Science & Technology 1999;33(2):262-269. |
R825406 (1998) R825406 (1999) R825406 (2000) R825406 (Final) |
not available |
|
Xia GS, Ball WP. Polanyi-based models for the competitive sorption of low-polarity organic contaminants on a natural sorbent. Environmental Science & Technology 2000;34(7):1246-1253. |
R825406 (2000) R825406 (Final) |
not available |
|
Zhang WX, Bouwer EJ, Ball WP. Bioavailability of hydrophobic organic contaminants: Effects and implications of sorption-related mass transfer on bioremediation. Ground Water Monitoring and Remediation 1998;18(1):126-138. |
R825406 (1997) R825406 (1998) R825406 (1999) R825406 (2000) R825406 (Final) |
not available |
Supplemental Keywords:
biodegradation, sorption, diffusion, intrinsic bioremediation, PAHs, polynuclear aromatic hydrocarbons, phenanthrene, naphthalene, remediation., Scientific Discipline, Toxics, Waste, Ecosystem Protection/Environmental Exposure & Risk, Bioavailability, National Recommended Water Quality, Environmental Chemistry, Chemistry, HAPS, Fate & Transport, Bioremediation, Ecological Risk Assessment, fate and transport, hydrocarbon, bioremediation model, Naphthalene, aquifer sediments, biodegradation, field studies, sorption kinetics, chemical speciation, saturated porous material, adsorption, chemical transport, kinetic studies, mass transfer, soils, toxicity, contaminants in soil, hazardous waste cleanup, soil characterization, saturated porous media, 1, 2-Dichlorobenzene, environmental toxicant, harmful environmental agents, mobility, aging, biodegradation of hydrophobic organic contaminants, contaminated aquifers, Phenanthrene, groundwater, hydrocarbon desorption kinetics, transportRelevant Websites:
DEPARTMENT OF Environmental Health and Engineering Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- 2000 Progress Report
- 1999 Progress Report
- 1998 Progress Report
- 1997 Progress Report
- Original Abstract
3 journal articles for this project