Published online 20 November 2007
Published in Vadose Zone J 6:1050-1056 (2007)
DOI: 10.2136/vzj2006.0157
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SPECIAL SECTION: HANFORD SITE
Experimental and Theoretical Assessment of the Lifetime of a Gaseous-Reduced Vadose Zone Permeable Reactive Barrier
E. C. Thorntona,
L. Zhonga,
M. Oostroma,* and
B. Dengb
a Environmental Technology Division, Pacific Northwest National Lab., Richland, WA 99352
b Dep. of Civil and Environmental Engineering, Univ. of Missouri, Columbia, MO 65211
* Corresponding author (mart.oostrom{at}pnl.gov).
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
Received 25 October 2006.
 |
ABSTRACT
|
|---|
The feasibility of using in situ gaseous reduction to establish a vadose zone permeable reactive barrier was evaluated through a combination of laboratory testing and consideration of fundamental vadose zone transport concepts. For the experimental evaluation, a series of laboratory column tests were conducted in which Hanford formation sediment from the USDOE Hanford Site in Richland, WA, was first treated with a diluted hydrogen sulfide gas mixture to reduce sediment iron oxide to ferrous sulfide. Water containing dissolved oxygen was then pumped through the columns at different flow rates to determine the reoxidation rate and the reductive capacity of the treated sediment. The results indicated that the treated sediment has a significant reductive capacity consistent with the basic reactions associated with the treatment and reoxidation processes. The observed reductive capacity was found to be dependent on the flow rate of water during the reoxidation phase of the tests. The reductive capacity approached the maximum value predicted on the basis of the treatment reaction as the flow rate was decreased. Thus, laboratory treatment tests provide a means for predicting the reductive capacity of the barrier under field conditions. In the theoretical assessment, oxygen diffusion was identified as the dominant mechanism leading to reoxidation of the barrier. Depending on vadose zone characterisitics, the predicted barrier lifetime varies from several years to more than 100 years.
Abbreviations: CPV, column pore volume
|
INTRODUCTION
|
|---|
Atheoretical approach was undertaken to estimate the lifetime of the vadose zone barrier at a site underlain by Hanford formation sediment from the USDOE Hanford Site in Richland, WA. An initial model assumed that the barrier lifetime is determined by the reoxidation of the barrier owing to the transport of oxygen through a vadose zone interval in which the moisture content of the sediment is moderately low. The results of this evaluation suggest that barrier reoxidation is primarily related to diffusion of oxygen through the gas-filled portion of the sediment pore space. If so, the barrier lifetime could be fairly short (several years). However, the presence of finer-grained strata with higher moisture content could potentially increase the barrier lifetime to greater than 100 years owing to a decrease in the effective diffusion coefficient for oxygen. Thus, detailed stratigraphic characterization and modeling information in conjunction with sediment treatment test results is needed to provide an accurate assessment of barrier lifetime at specific sites.
In situ gaseous treatment of sediments with diluted hydrogen sulfide is a potentially effective method for immobilizing selected toxic metals and radionuclides in the vadose zone because of the lower solubility of the reduced contaminant or the formation of insoluble sulfide compounds. The mobility of contaminant constituents in the vadose zone is an important environmental concern primarily because transport to underlying aquifers can result in significant groundwater contamination. The USDOE's Hanford Site in Washington State, for example, has a thick vadose zone interval containing contamination that has impacted groundwater quality (Riley et al., 1992).
A significant amount of research to date indicates the potential for contaminant immobilization through gaseous treatment with hydrogen sulfide (H2S). Chromate (hexavalent chromium), in particular, is readily reduced by hydrogen sulfide to the trivalent oxidation state, which is relatively insoluble and generally not readily reoxidized to the hexavalent state (Thornton and Amonette, 1999; Thornton, 2000; Thornton et al., 1999). Uranium and technetium can also be reduced and immobilized by hydrogen sulfide treatment (Thornton et al., 2003). Hua et al. (2006) have recently shown that the rate of U(VI) reduction by H2S depends strongly on solution pH and carbonate concentrations. Unlike chromium, uranium and technetium are potentially susceptible to reoxidation and possible remobilization once the environment returns to oxidizing conditions. Testing results presented by Zhong et al. (2007), however, indicate that remobilization of uranium is inhibited by adsorption to sediment reaction products.
In addition to direct reduction of contaminant species, gaseous treatment could be used to generate a chemically reduced permeable reactive barrier in the vadose zone through the reduction of sediment ferric iron oxide phases. The reduced ferrous iron could thus potentially immobilize contaminants by reduction and maintain a reduced environment that would prevent remobilization of these contaminants for a substantial period of time. Laboratory tests have consequently been undertaken to provide information related to aspects of sediment reduction and contaminant immobilization by gaseous treatment (Cantrell et al., 2003; Thornton et al., 2006; Zhong et al., 2007). These studies have confirmed that interaction of gaseous hydrogen sulfide reduces ferric oxides and hydroxides in the absence of oxygen to generate ferrous sulfide. The stochiometry of the reaction provides a basis for estimating the reductive capacity of treated sediment if the amount of hydrogen sulfide consumed during treatment is measured, or if the amount of reactive ferric hydroxide in a sediment sample is known. This information is important for designing the vadose zone barrier and for predicting the effective barrier lifetime.
The primary objective of the experiments reported here was to determine the reductive capacity generated by treatment of Hanford formation sediment with diluted hydrogen sulfide gas. A series of column tests was performed in which a Hanford formation soil sample was treated with diluted hydrogen sulfide gas and the subsequent reductive capacity of the treated soil measured through reoxidation of the soil with oxygenated water. This approach is similar to that used by Szecsody et al. (1998) to evaluate the reductive capacity of dithonite-treated sediments and the potential to immobilize uranium within the Hanford unconfined aquifer. The oxygenated water was pumped through the treated soil in the current study at several flow rates (pore velocities) to determine if the sediment reoxidation process is rate limited and, if so, to determine the relationship between pore velocity and reoxidation rate.
A theoretical evaluation is also presented that relies on analytical solutions to define the oxygen flux terms in the vadose zone that control eventual reoxidation of the H2S-reduced barrier. This model is based on equations presented by Jury and Horton (2004) and considers both diffusional and advectional fluxes as oxygen gas diffuses downward to the barrier and as oxygenated water infiltrates through it. The oxygen flux estimate and the measured reductive capacity of the treated sediment is then used to predict the lifetime of the barrier based on the time required to reoxidize the treated zone.
|
Materials and Methods
|
|---|
Sediment Characterization and Preparation
The sediment used in these tests was uncontaminated Hanford formation sand collected in the 200 East Area of the Hanford Site, Richland, WA. The sediment was air-dried and sieved to limit soil particle sizes to <2 mm before testing.
Characterization information for the sediment has been previously presented in Thornton et al. (2006). In summary, the sample used in testing is a medium- to coarse-grained sand. X-ray diffraction analysis of the Hanford formation bulk fraction by Serne et al. (2002) indicates that it is composed of quartz (30–50%), plagioclase feldspar (5– 20%), and a minor amount of potassium feldspar (<10%). Serne et al. (2002) also reported that the total iron content of four Hanford formation samples ranged between 3.46 to 9.64% w/w as Fe2O3. Much of this iron is likely associated with basaltic rock fragments that typically comprise a significant portion of Hanford formation sediments.
In addition, the amorphous iron oxide present in the sediment was extracted with 0.5 M HCl in the current study as described by Heron et al. (1994). Colorimetric analysis of the extract performed by the phenathroline method (Loeppert and Inskeep 1996) indicated that 0.29% w/w total iron (0.0029 g total iron per gram soil) is present in the sediment as amorphous iron oxide.
Column Tests
Air-dried uncontaminated sediment was packed into columns 2.6 cm in diameter and 30.4 cm in length. The porosity and column pore volume (CPV) were calculated for each of four separate packings based on the column dimensions, sample weight, residual sediment moisture content (
0.015 g water per gram soil), and an estimated particle density (2.7 g cm–3; Serne et al., 2002). Column parameters for the four tests performed are presented in Table 1. Calculated sample porosity in the four column tests ranged from 0.307 to 0.322.
The sediment columns were treated with diluted dry hydrogen sulfide gas during each test (Fig. 1
). Gas sources included a small cylinder of 1% hydrogen sulfide in nitrogen and a full-size cylinder of nitrogen for dilution. A 200 mg L–1 hydrogen sulfide mixture was prepared with flow controllers, and the mixture was directed downward through the column during treatment at a flow rate of 500 mL min–1. Inflow and effluent concentrations (Co and C, respectively) of hydrogen sulfide were monitored to determine treatment progress using hydrogen sulfide sensors (EIT Corp., Sterling, VA). Treatment was terminated in the column tests once C/Co reached approximately 0.9, equivalent to an effluent concentration, C, of 180 mg L–1 H2S with an influent concentration, Co, of 200 mg L–1.
An aerated solution was prepared for reoxidation of the treated sediment by bubbling a mixture of 21.5% oxygen in nitrogen through deionized water. This solution was pumped upward through the column using a syringe pump (Kloen Co. Ltd., Summerlin, NV) during the four tests conducted (Fig. 2
). The reoxidation stage of the tests was performed at four flow rates, ranging from 0.096 to 0.71 mL min–1, to obtain information regarding the amount of oxygen consumed as a function of column residence time and pore velocity (Table 1). The dissolved oxygen content of the column influent and effluent solutions was monitored with a calibrated oxygen electrode (Microelectrodes, Inc., Bedford, NH) and recorded using a LabView data acquisition system (National Instruments, Austin, TX). Reoxidation of the treated sediment was considered complete when a C/Co value of >0.75 was achieved in terms of the dissolved oxygen content of the column effluent relative to the influent.
|
Results and Discussion
|
|---|
Column Test Results
The treatment H2S breakthrough curves were all similar to that shown in Fig. 3
for test R-1. Initial hydrogen sulfide breakthrough occurred within several hours of test initiation. The curves rose sharply until a C/Co breakthrough ratio of about 0.5 was achieved and then began to flatten out. The breakthrough ratio of 0.5 was achieved in the treatment of the four tests at 3.47 ± 0.22 h, indicating good reproducibility for the treatment process. About 32 h (range: 31.33–32.45 h) were required to complete treatment for the four tests as defined by attainment of a C/Co hydrogen sulfide value of approximately 0.9.
The consumption of hydrogen sulfide during treatment can be obtained by the breakthrough curve data (Fig. 3) in conjunction with the associated flow rate. This information can then be used to estimate the amount of Fe(II) generated by reduction of Fe(III) through the following reaction stoichiometry:
 | [1] |
The amount of ferrous iron generated was 0.00026 g Fe(II) per gram soil during the treatment phase of the four tests based on the above reaction and the amount of H2S consumed. Analysis of the sediment used in this study indicated that 0.0029 g iron per gram soil was present as determined by 0.5 M HCl extraction, which is roughly equal to the amount of reactive ferric iron available in the soil for reduction by hydrogen sulfide (Cantrell et al., 2003). Thus, only about 9% of the reactive Fe(III) component was reduced in the tests performed in this study. This appears to be primarily a result of the relatively short treatment time involved (32 h). Previous treatment tests indicate that the calculated conversion of available Fe(III) to Fe(II) has been as much as 56% when longer treatment times have been used (Thornton et al., 2006). Gaseous reduction of sediment can also be enhanced by using humidified treatment gas (Zhong et al., 2007).
The reoxidation curves for the four column tests are presented in Fig. 4
for the first 60 CPVs. The number of CPVs of water associated with initial oxygen breakthrough (i.e., C/Co > 0) was observed to decrease at higher flow rates. This indicates that the reoxidation reaction is rate limited; that is, more oxygen is consumed per a given number of CPVs at lower flow rates (i.e., greater column residence time). A second stage of lower oxygen consumption followed the initial breakthrough of oxygen, which continued throughout the duration of the tests (Fig. 4). Longer-term consumption of oxygen continues after initial oxygen breakthrough, but immobilization of contamination by a subsurface reductive barrier is less likely to occur after the system starts to become aerobic.
View larger version (24K):
[in this window]
[in a new window]
|
FIG. 4. Oxygen breakthrough curves for the reoxidation tests during the first 60 column pore volumes. (Q, volumetric flow rate.)
|
|
The reaction involved in the reoxidation step of the tests may be written as
 | [2] |
where it is assumed that elemental sulfur is the predominant oxidized sulfur product (Cantrell et al., 2003; Kim et al., 2001). This reaction provides a basis for predicting the maximum amount of oxygen that could be consumed during reoxidation given the amount of FeS generated during treatment. As indicated above, the amount of ferrous iron generated was approximately 0.00026 g Fe(II) per gram soil according to Eq. [1], which could consume a maximum of 0.00011 g oxygen per gram soil per Eq. [2].
The cumulative mass of oxygen consumed during the reoxidation tests is presented in Fig. 5
as a function of the number of CPVs of aerated water pumped. Figure 5 illustrates that the cumulative amount of oxygen consumed during the tests ranged from 41.5 to 58.3% of the maximum potential oxygen consumption. The figure also indicates that the degree of consumption of oxygen was somewhat greater at the lower flow rates. The highest rate of oxygen consumption was associated with the first 20 CPVs of oxygenated water pumped through the columns during this study, which is the period of treatment before the initial breakthrough of oxygen (compare Fig. 4 and Fig. 5).
View larger version (23K):
[in this window]
[in a new window]
|
FIG. 5. Relationship between cumulative oxygen consumed and the number of pore volumes of water pumped through the columns during the reoxidation tests. (Q, volumetric flow rate.)
|
|
The linear pore velocity, v, of aerated water during the reoxidation step of each of the column tests can be calculated from the following relationship:
 | [3] |
where Q is the volumetric flow rate (ml min–1), 
is porosity, and A is the cross-sectional area of the column (cm2). Values of pore velocities for the column tests conducted in this study are presented in Table 1. Figure 6
presents the relationship between linear pore velocity and the cumulative amount of oxygen consumed in the column tests at the time of initial breakthrough of oxygen. A best fit between the consumption of oxygen and pore velocity was determined using a function of the natural logarithm of the pore velocity. The figure illustrates that the total amount of oxygen consumed increases significantly as pore velocity decreases. This suggests that at the low flow rates associated with normal infiltration, most of the potential oxygen consumption capacity in treated soil would be available during reoxidation of the soil before the system starts to become aerobic (i.e., initial oxygen breakthrough occurs).
View larger version (14K):
[in this window]
[in a new window]
|
FIG. 6. Relationship between total oxygen consumed during initial oxygen breakthrough and pore velocity in the reoxidation tests.
|
|
The results of the reoxidation tests indicate that a significant reduction capacity can be achieved by treatment of Hanford formation sediments with hydrogen sulfide gas mixtures. This information provides a basis for initial predictions of barrier lifetime, as illustrated below. Future work should be undertaken, however, for longer treatment durations and at lower flow rates during reoxidation to support full-scale deployment of the gaseous treatment approach. Larger-scale laboratory and small- to intermediate-scale field tests should be considered as a means for obtaining this information.
Estimated Lifetime of a Vadose Zone Barrier Generated by Gaseous Treatment
The results of this study can be used to estimate the lifetime of a vadose zone barrier generated by in situ gaseous reduction. The approach presented here assumes that the lifetime of the barrier is related to reoxidation of iron sulfide within the treated zone of the Hanford formation in response to introduction of oxygen to the barrier. The rate of transport of oxygen in the vadose zone is based on equations previously presented by Jury and Horton (2004).
The total oxygen flux consists of four components: diffusion related to movement of oxygen from the surface to the treated vadose zone interval through the gas-filled portion of the sediment pore space (JgO2), advective transport of oxygen dissolved in the aqueous phase within the pore space (FlO2), diffusion through the water-filled portion of the pore space (JlO2), and advective transport or convection of oxygen gas downward through the vadose zone (FgO2). The total oxygen flux to the vadose zone barrier (
O2,flux) is given by
 | [4] |
The first flux term,JgO2, refers to oxygen transported from the surface (z = 0) to the top of the reduced sediment barrier (z = L) by diffusion through the gas-filled portion of the pore space only. The mass balance equation for vertical, one-dimensional diffusive oxygen transport without reaction can be written as
 | [5] |
where t is time (T), z is the vertical distance to the top of the barrier (L), CgO2 is the oxygen concentration in the gas phase (M L–3) and JgO2 is the diffusive mass flux of oxygen in the gas phase (M L–2 T–1). The diffusive mass flux term is defined as
 | [6] |
where sg is gas saturation (i.e., fractional portion of the diffusive porosity that is gas-filled), nD is diffusive porosity, 
g is the gas tortuosity factor, and DgO2 is the binary gaseous diffusion coefficient for oxygen in air (L2 T–1).
Inserting Eq. [6] into Eq. [5] and assuming that gas saturation, diffusive porosity, tortuosity, and the diffusion coefficient are constant for 0 
z L yields
 | [7] |
Assume furthermore that the diffusional transport process is at steady state (i.e., 
CgO2/t=0). Equation [7] reduces to
 | [8] |
This simple second order ordinary differential equation can be integrated twice to yield
 | [9] |
where 1 and 2 are the integration constants.
Assuming that the oxygen concentration is zero at the top of the barrier (i.e., the reduction rate is fast within the treated zone), the boundary conditions appropriate for this problem are CgO2=Co at z = 0 and CgO2=0 at z = L. Equation [9] can then be solved to yield
 | [10] |
Equation [10] indicates that for diffusive transport from the surface to the top of the barrier, given the assumptions identified above, the oxygen concentration profile will be linear (i.e., dCgO2/dz=–Co/L). The linearity of the concentration profile results in a constant oxygen diffusive flux. To reflect this, Eq. [6] can be rewritten as
 | [11] |
and the diffusive molar flux (mol L–2 T–1) can be computed using
 | [12] |
where MO2 is the molecular weight (M mol–1) of oxygen (0.032 kg mol–1).
Tortuosity, which is used to calculate diffusive flux in porous media, can be estimated by Millington and Quirk's (1959) model:
 | [13] |
where 
= l for the aqueous phase or g for the gas phase. If we assume sg = 0.9 and nD = 0.2, for example, then g = 0.457.
The oxygen diffusion coefficient can be computed through several formulations. The Wilke and Lee method (Reid et al., 1987) is widely used for diffusion coefficients of gas components. For a temperature of 15°C, the diffusion coefficient is 0.21 cm2 s–1 or 2.1 x 10–5 m2 s–1.
The oxygen concentration at the surface can be computed using the ideal gas law. Assuming a gas mixture containing 21% oxygen, Co = 0.28 kg m–3.
Using the values given above, Eq. [12] yields a relationship for calculating the diffusive molar flux:
 | [14] |
For a depth to the top of the barrier of L = 25 m, for example, the diffusive oxygen molar flux would be about 19 mol m–2 yr–1.
The second component of oxygen flux, FlO2, is associated with the advective transport of oxygen via downward flow or infiltration of the aqueous phase within the pore space. If we assume that the dissolved concentration of oxygen in the vadose zone is constant, then
 | [15] |
where ClO2 is the oxygen concentration in the aqueous phase (M L–3) and qz is the specific discharge or infiltration rate associated with aqueous phase flow (L T–1).
Assuming a moderately low infiltration rate of qz = 0.1 m yr–1 and an oxidized vadose interval above the barrier with ClO2=8 mg L–1 or 0.008 kg m–3, the advective oxygen molar flux associated with these values will only be 0.025 mol m–2 yr–1. Thus, the oxygen advective flux term FlO2 is much less significant than the diffusional flux term JgO2 from the standpoint of barrier reoxidation.
Considering next JlO2, the flux term related to diffusion of oxygen through the portion of the pore space occupied by water, and following the logic presented above for diffusional flux,
 | [16] |
where sl = 0.1, nD = 0.2, DlO2=2x10–9 m2 s–1, and Co=0.008 kg m–3. The value for tortuosity within water-filled portion of the pore space, l, calculated using Eq. [13] and the values presented above, is 0.00272. For these values, Jl,MO2 is equal to only 3.4 x 10–8 mol m–2 yr–1 and thus is insignificant.
The final flux term, FgO2, is difficult to estimate. This term represents convection of gas in the vadose zone and can include temperature, barometric pressure, wind, and rainfall effects (Jury and Horton, 2004). Generally these are near-surface soil phenomenon. Thus, FgO2 should also be much less than JgO2.
The lifetime of the barrier will depend on the oxygen flux in the vadose zone and the reductive capacity of the treated sediment. The reductive capacity of the barrier is related to the amount of FeS or ferrous iron generated during gaseous treatment. As discussed above, the laboratory treatment and reoxidation study suggests that the treated zone may remain anaerobic as long as some FeS remains, provided the infiltration rate is low. Also, it is judged that at least 50% of the available iron in the vadose zone may be converted to FeS if the treatment time is sufficiently long. Thus, for the sediment tested, let us assume that 0.0015 g of the 0.0029 g of available iron per gram of sediment is converted to FeS and the treated interval remains anaerobic until all of the FeS is reoxidized. For a dry bulk sediment density of 1850 kg m–3, and given the molecular weight of iron (0.0558 kg mol–1), the maximum molar concentration of ferrous iron that could be generated by gaseous reduction is about 50 mol m–3. The stoichiometry of the reoxidation process, as presented above by Eq. [2], indicates that 0.75 moles of oxygen are required to completely reoxidize 1 mole of FeS. Thus, about 37.5 moles of oxygen will oxidize a cubic meter of the gas-treated soil.
Given this estimate of the oxygen-consuming capacity (O2,capacity) of the reduced sediment and the assumption that the major oxygen flux term is JgO2, we can now determine the time required to reoxidize a cubic meter of the barrier. Given the barrier thickness T in meters, the amount of time required, tr, in years, can be computed from the following expression:
 | [17] |
A lifetime of about 2 yr is expected for a barrier with unit thickness (1 m) if the total oxygen flux term is roughly equal to JgO2. Thus, the barrier would undergo reoxidation in a relatively short period of time based on this model. Fresh injections of treatment gas could potentially be undertaken periodically, however, to extend the lifetime of the barrier if this situation is found to exist.
The flux of oxygen will decrease substantially if the moisture content is higher (Huesemann and Truex, 1996). This will result in an increase in the lifetime of the barrier. If we assume, for example, that sg = 0.7, then the calculated barrier lifetime will increase to about 4.5 yr. It is unlikely, moreover, that the assumption of a homogeneous vadose zone is valid. In particular, the occurrence of finer-grained strata with higher moisture contents could reduce the flux of oxygen considerably and thus result in a substantially longer barrier lifetime.
It is also possible that the Millington and Quirk (1959) model may not be valid at high moisture content values. Williams and Oostrom (2000) observed experimentally that reoxygenation in a fluctuating water table system was considerably less than predicted by modeling results. They were able to obtain an adequate match between laboratory data and modeling results by lowering the gas-phase molecular diffusion coefficient of oxygen. This was based on the work of Nielson et al. (1984), who proposed that the effective oxygen diffusion coefficient in soils can vary greatly at higher moisture contents depending on the texture of the sediment. In particular, for poorly sorted fine-grained sediments, a film of water blocking some of the pore spaces limits the diffusive flux. In effect, oxygen is diffusing through both the aqueous and gas phases within the pore space when partial blockage occurs. The effective diffusion coefficient for oxygen will thus be a value intermediate between the diffusional coefficients for the gas and aqueous phases for zones where a substantial amount of pore blockage occurs.
The situation in which all of the pore spaces in the vadose zone are partially blocked and oxygen diffusion occurs in both the aqueous and gas phases illustrates the pore blockage effect. The diffusive molar flux, based on an effective diffusion coefficient Je,MO2, is
 | [18] |
Assuming that the effective tortuosity, e, is equal to 0.5 and that sl = sg = 0.5, the effective saturation, se, will be equal to 1, since we are considering both the aqueous and gas phases within the pore spaces. As before, nD is equal to 0.2. Let Co be equal to 0.14 kg m–3, which is the average of the oxygen concentrations of the gas and aqueous phases in equilibrium with the atmosphere. Let us also assume that the effective oxygen diffusion coefficient is 5 x 10–7 m2 s–1, which is a value intermediate between the aqueous phase diffusion coefficient (2 x 10–9 m2 s–1) and the gas phase diffusion coefficient (2 x 10–5 m2 s–1) (Williams and Oostrom, 2000). Based on these values, Je,MO2 is equal to 0.28 mol m–2 yr–1 if the top of the barrier is located at a depth of 25 m below the surface. Using this flux value, Eq. [17] indicates a barrier lifetime of about 135 yr.
It is clearly not possible to estimate accurately the lifetime of a vadose zone barrier without sufficiently detailed information regarding the geohydrologic characteristics of the interval between the ground surface and the top of the barrier. In particular, some degree of heterogeneity is generally present in the vadose zone. For example, a combination of coarse-grained, unsaturated and finer-grained, partially saturated strata may be present at a specific site. Depending on the characteristics of the vadose zone, the barrier lifetime could vary from several years to more than 100 years. Thus, a reasonably accurate prediction of barrier lifetime must be based on a model that incorporates estimates of oxygen flux changes resulting from stratigraphic changes in the vadose zone.
|
Conclusions
|
|---|
Results of laboratory column tests performed in this study indicate that sediment treated with diluted hydrogen sulfide has a significant reducing capacity. The total reductive capacity of the treated sediment is related to the amount of FeS generated during treatment. The measured reductive capacity is dependent on the flow rate of oxygenated water, owing to kinetic limitations, but appears to approach an upper limiting value at lower flow rates that depends on the amount of FeS generated. Thus, reoxidation tests conducted in the laboratory at low flow rates should serve as a means for estimating the total reductive capacity of treated sediment under field conditions.
Theoretical considerations indicate that diffusion of oxygen through the vadose zone to the top of the barrier will be the dominant mechanism leading to reoxidation of the barrier. The lifetime could be as short as several years if the sediments in the vadose zone are fairly coarse grained and unsaturated. It is likely that the lifetime could be much longer, however, if the vadose zone is heterogeneous in nature with intervals of finer-grained, partially saturated strata. Accurate prediction of barrier lifetime for specific sites will therefore require detailed characterization of the vadose zone and formulation of models that adequately incorporate this heterogeneity. Additional experimental studies may also be required to better understand the diffusion of oxygen in partially saturated sediments where a combination of transport through mixed aqueous and gas- filled pore space is involved. Field investigations are also needed that will provide verification of the anticipated treatment and reoxidation processes and rates under full-scale conditions.
|
ACKNOWLEDGMENTS
|
|---|
Dr. Jim Szecsody of Pacific Northwest National Laboratory (PNNL) helped set up the data acquisition system, and Dr. Kirk Cantrell (PNNL) developed an initial version of the theoretical model for estimation of barrier lifetime. We would also like to thank Mr. Mike Truex (PNNL) who provided constructive review comments during the preparation of the manuscript. The funding for this study was provided by the Environmental Management Science Program (EMSP) of the USDOE (Grant No. DE-FG02-03ER63616). Pacific Northwest National Laboratory is operated by Battelle for the USDOE. The effort by M. Oostrom was supported by the Environmental Molecular Sciences Laboratory, a national scientific user facility.
|
REFERENCES
|
|---|
- Cantrell, K.J., S.B. Yabusaki, M.H. Engelhard, A.V. Mitroshkov, and E.C. Thornton. 2003. Oxidation of H2S by iron oxides in unsaturated conditions. Environ. Sci. Technol. 37:2192–2199.[Medline]
- Heron, G., C. Crouzet, A.C.M. Bourg, and T.H. Christensen. 1994. Speciation of Fe(II) and Fe(III) in contaminated aquifer sediments using chemical extraction techniques. Environ. Sci. Technol. 28:1698–1705.
- Hua, B., H. Xu, J. Terry, and B. Deng. 2006. Kinetics of uranium(VI) reduction by hydrogen sulfide in anoxic aqueous systems. Environ. Sci. Technol. 40:4666–4671.[Medline]
- Huesemann, M.H., and M.J. Truex. 1996. The role of oxygen diffusion in passive remediation of petroleum contaminated soils. J. Hazard. Mater. 51:93–113.[CrossRef][Web of Science]
- Jury, W.A., and R. Horton. 2004. Soil physics. John Wiley & Sons, Hoboken, NJ.
- Kim, C., Q. Zhou, B. Deng, E.C. Thornton, and H. Xu. 2001. Chromium(VI) reduction by hydrogen sulfide in aqueous media: Stoichiometry and kinetics. Environ. Sci. Technol. 35:2219–2225.[Medline]
- Loeppert, R.L., and W.P. Inskeep. 1996. Iron. p. 639–664. In D.L. Sparks (ed.) Methods of soil analysis: Part 3. Chemical methods. SSSA Book Ser. 5. ASA and SSSA, Madison, WI.
- Millington, R.J., and J.P. Quirk. 1959. Permeability of porous media. Nature 183:387–388.[CrossRef]
- Nielson, K.K., V.C. Rogers, and G.W. Gee. 1984. Diffusion of radon through soils: A pore distribution model. Soil Sci. Soc. Am. J. 4:482–487.
- Reid, R.C., J.M. Prausnitz, and B.E. Poling. 1987. The properties of gases and liquids. McGraw-Hill, New York.
- Riley, R.G., J.M. Zachara, and F.J. Wobber. 1992. Chemical contaminants on DOE lands and selection of contaminant mixtures for subsurface science research. DOE/ER-0547T. USDOE, Office of Energy Research, Washington, DC.
- Serne, R.J., B.N. Bjornstad, H.T. Schaef, B.A. Williams, D.C. Lanigan, D.G. Horton, R.E. Clayton, A.V. Mitroshkov, V.L. Legore, M.J. O'Hara, C.F. Brown, K.E. Parker, I.V. Kutnyakov, J.N. Serne, G.V. Last, S.C. Smith, C.W. Lindenmeier, J.M. Zachara, and D.S. Burke. 2002. Characterization of vadose zone sediment: Uncontaminated RCRA borehole core samples and composite samples. PNNL-13757-1. Pacific Northwest National Laboratory, Richland, WA.
- Szecsody, J.E., K.J. Cantrell, K.M. Krupka, C.T. Resch, M.D. Williams, and J.S. Fruchter. 1998. Uranium mobility during in situ redox manipulation of the 100 Areas of the Hanford Site. PNL-12048. Pacific Northwest Laboratory, Richland, WA.
- Thornton, E.C. 2000. In situ gaseous reduction. p. 1302–1307. In B.B. Looney and R.W. Falta (ed.) Vadose zone science and technology solutions. Battelle Press, Columbus, OH.
- Thornton, E.C., and J.E. Amonette. 1999. Hydrogen sulfide gas treatment of Cr(VI)-contaminated sediment samples from a plating-waste disposal site: Implications for in-situ remediation. Environ. Sci. Technol. 33:4096–4101.
- Thornton, E.C., J.T. Giblin, T.J. Gilmore, K.B. Olsen, J.M. Phelan, and R.D. Miller. 1999. In situ gaseous reduction pilot demonstration: Final report. PNNL-12121. Pacific Northwest National Laboratory, Richland, WA.
- Thornton, E.C., V.L. Gore, and K.B. Olsen. 2003. Immobilization of chromium, technetium, and uranium by gaseous reduction in soil from the Hanford Site. p. 3.3-10–3.3-14. In M.J. Hartman, L.F. Morasch, and W.D. Webber (ed.) Hanford Site groundwater monitoring for fiscal year 2002. PNNL-14187. Pacific Northwest National Laboratory, Richland, WA.
- Thornton, E.C., L. Zhong, and M. Oostrom. 2006. Development of a field design for in situ gaseous treatment of sediment based on laboratory column test data. J. Environ. Eng. 132:1626–1632.[CrossRef]
- Williams, M., and M. Oostrom. 2000. Oxygenation of anoxic water in a fluctuating water table system: An experimental and numerical study. J. Hydrol. 230:70–85.[CrossRef]
- Zhong, L., E.C. Thornton, and B. Deng. 2007. Uranium immobilization by hydrogen sulfide gaseous treatment under vadose zone conditions. Vadose Zone J. 6:149–157.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
G. W. Gee, M. Oostrom, M. D. Freshley, M. L. Rockhold, and J. M. Zachara
Hanford Site Vadose Zone Studies: An Overview
Vadose Zone J.,
November 20, 2007;
6(4):
899 - 905.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
