Xu, X., N.R. Thomson, 2010. Hydrogen peroxide persistence in the presence of aquifer materials. Soil Sed. Contam. 19(5). xx-xx.

In situ chemical oxidation using hydrogen peroxide for groundwater remediation has received a great deal of attention over the past decade.  A comprehensive investigation that focused on the interaction between hydrogen peroxide and eight aquifer materials was performed using both batch and column experiments.  The results from the batch experiments indicated that the decomposition of hydrogen peroxide in the presence of various aquifer materials followed a first-order rate law, and was strongly affected by the content of amorphous transition metals.  The exposure of the aquifer solids to hydrogen peroxide for 14 days indicated that not all forms of NOM contributed the observed decomposition.  Multiple linear regression analysis was used to generate two potentially useful predictive decomposition rate coefficient relationships based on the various aquifers material characteristics.  Column experiments were conducted with five representative aquifer materials to complement and expand the findings from the batch experiments.  As expected the decomposition rate coefficients were higher in the column experiments relative to the batch experiments due to the larger solids mass to solution volume ratio.  Attempts to simulate the column observations indicated that dissolution and transport of metals from aquifer solids may play an important role in hydrogen peroxide persistence in some subsurface environments.

Xu, X., N.R. Thomson, 2009. A long-term bench-scale investigation of permanganate consumption by aquifer materials. J. Contam. Hydrol. doi:10.1016/j.jconhyd.2009.09.001.

In situ chemical oxidation (ISCO) applications using permanganate involve the injection or release of permanganate into the subsurface to destroy various target contaminants. Naturally occurring reduced components associated with aquifer materials can exert a significant oxidant demand thereby reducing the amount of permanganate available for the destruction of contaminants as well as reducing the overall rate of oxidation. Quantification of this natural oxidant demand (NOD) is a requirement for site-specific assessment and the design of cost-effective oxidant delivery systems. To further our understanding of the interaction between permanganate and aquifer materials, aerobic and anaerobic aquifer materials from eight representative sites throughout North America were tested in a series of systematic bench-scale experiments. Various permanganate to aquifer solids mass loading ratios at different initial permanganate concentrations in well-mixed batch reactors were monitored for >300 days. All NOD temporal profiles demonstrated an initial fast consumption rate followed by a persistent slower consumption rate. The data generated show that the mass loading ratio, the initial permanganate concentration, and the nature and quantity of reduced aquifer material species are the main factors controlling permanganate consumption rates. A higher initial permanganate concentration or a larger mass loading ratio produced a larger fast NOD consumption rate and generated a corresponding higher maximum NOD value Hence, both the NOD temporal profile and the maximum NOD are not single-valued but are heavily dependent on the experimental conditions. Predictive relationships were developed to estimate the maximum NOD and the NOD at 7 days based on aquifer material properties. The concentration of manganese oxides deposited on the aquifer solids was highly correlated with the mass of permanganate consumed suggesting that passivation of the NOD reaction sites occurred due to the formation of manganese oxide coating on the grains. A long-term NOD kinetic model was developed assuming a single fast and slow reacting oxidizable aquifer material species, passivation of NOD reaction sites, and the presence of an autocatalytic reaction. The developed model was able to successful capture the observed NOD temporal profiles, and can be used to estimate in situ NOD behavior using batch reactor experimental data. The use of batch tests to provide data representative of in situ conditions should be used with caution.

Lambert, J., T. Yang, N.R. Thomson, J.F. Barker. 2009. Pulsed biosparging of a residual fuel source emplaced at CFB Borden, International Journal of Soil, Sediment and Water. In press.

Biosparging enhances both aerobic biodegradation and volatilization, and is commonly applied to residual hydrocarbon source zone remediation. This technology was applied in pulsed mode to a known source of gasoline contaminatioll in order to quantify the extent of remediation achieved in terms of both mass removed and reduction in mass discharge into groundwater. The gasoline source was created by injecting about 1.0 L of gasoline with 10% ethanol in small volumes from 24 injection points below the water table in 2004. The downgradient plume is still being monitored and the source area was cored in 2007. In 2008, a 3-point biosparge system was operated with an airtight cover to capture and monitor off-gases. Of the hydrocarbons in place, about 80% of pentane, 50% of hexane, but only about 4% of the aromatic hydrocarbons were volatilized and removed. CO2 and O2 monitoring of the off-gas confirmed limited biodegadation of hyclrocarbons.

Javor, P., N.R. Thomson, B. Wotton, 2009. GeoHalifax, Proceeding from the 62nd Canadian Geotechnical Conference and the 10th Joint CGS/IAH-CNC Groundwater Conference, Halifax, NS, Sept. 20-23.

Large diameter residential drinking water wells are at a higher risk of contamination from surface water impacts than drilled wells.  The possibility of a higher incidence of contamination of large diameter wells is attributed to site selection and construction problems. The objective of this investigation is to assess several design changes that are thought to improve the structural integrity of large diameter wells and to determine whether one design is more prone to contamination than the others.  This paper describes the construction of the large diameter wells and results from ongoing laboratory and field trials.

Sousa, M.R., D.L. Rudolph, N.R. Thomson, B. Conant Jr, 2009. Predicting Effects of Agricultural Beneficial Management Practices on Water Supply Wells using Numerical Modelling, Proceeding from the 62nd Canadian Geotechnical Conference and the 10th Joint CGS/IAH-CNC Groundwater Conference, Halifax, NS, Sept. 20-23.

Agricultural Beneficial Management Practices (BMPs) were implemented within the capture zone of water supply wells at a field site in southern Ontario, Canada, in an attempt to reduce nitrate concentrations in wells impacted by non-point source contamination. A 3-dimensional, variably-saturated, flow and transport model (FEFLOW) was subsequently applied and predicted the effects of current and future BMP scenarios on the water quality. Simulations indicate reductions in nutrient applications can reduce nitrate concentrations in the supply wells.  Complex glacial deposits and large depths to water influenced the timing and magnitude of the simulated improvements.

O’Connell, M., N.R. Thomson, K. Sra, C. Bright, T. Treeratanaphitak, 2009. Optimizing solvent extraction and chemical oxidation of PCBs.  Proceedings from the Fifth International Conference on Remediation of Contaminated Sediments, Jacksonville FL, Feb. 2-5.

Polychlorinated biphenyls (PCBs) are carcinogenic persistent contaminants and although their manufacturing in North America ceased in the late 1970s, their high heat resistance made their use widespread over their production lifetime. PCB contaminated soil on industrial and other sites has historically been dealt with through excavation followed by off-site disposal or incineration. Companies have successfully practiced various solvent extraction technologies for PCB separation, primarily in the southern United States. Although successful at removing a large quantity of PCBs from soil, this technology can be improved upon by making the extraction more complete, efficient and suitable for use in a variety of climates. Laboratory and field research underway at the University of Waterloo will identify the factors controlling PCB extraction with solvents in order to optimize PCB extraction as it is applied on different soil types and in various climates. The collected data will be used to develop a temperature-corrected kinetic model to better represent the extraction process. Research will also explore PCB contaminated soil pretreatment using chemical oxidation. As past research has shown that weathered PCB in soil is more difficult to remove, contaminated field samples from Southern Ontario, Canada are being used for this work, rather than synthetically prepared samples. Initial screening experiments identified optimal soil moisture contents for different extraction solvents. Additional experiments were conducted to discover if the influence of factors controlling PCB extraction was temperature dependent. Experiments were conducted to assess the treatability of oxidizing PCBs with persulfate and peroxide. Experiments are currently being run to assess the combined use of polar and non-polar solvents. This presentation will include a discussion of the observed trends, experiments underway, factors influencing solvent extraction, and the expected form of the kinetic model.

Nelson, L., J. Barker, T. Li , N. Thomson, M. Ioannidis, J. Chatzis, 2009. A field trial to assess the performance of CO2-supersaturated water injection for residual volatile LNAPL recovery. J. Contam. Hydrol. 109, 82-90. doi:10.1016/j.jconhyd.2009.08.002.

A pilot scale field trial was conducted to evaluate the recovery of volatile, light non-aqueous phase liquids (LNAPLs) using a novel remediation method termed supersaturated water injection (SWI). SWI uses a patented technology to efficiently dissolve high concentrations of CO2 into water at elevated pressures. This water is injected into the subsurface resulting in the nucleation of CO2 bubbles at and away from the injection point. The nucleating bubbles coalesce, rise and volatilize residual LNAPL ganglia. In this study, an LNAPL composed of 103 kg of volatile pentane and hexane, and 30 kg of non-volatile Soltrol was emplaced below the water table at residual saturation. The SWI technology removed 78% of the pentane and 50% of the less volatile hexane. Contaminant mass was still being removed when the system was shut down for practical reasons. The mass removed is comparable to that expected for air sparging but a much smaller volume of gas was injected using the SWI system.

Thomson, N.R., M. Fraser, C. Lamarche, J.F. Barker, S. Forsey, 2008. Rebound of a creosote plume following partial source zone treatment with permanganate. J. Contam. Hydrol. 102, 154-171. doi: 10.1016/j.jconhyd.2008.07.001.

The long-term management of dissolved plumes originating from a coal tar creosote source is a technical challenge.  For some sites stabilization of the source may be the best practical solution to decrease the contaminant mass loading to the plume and associated off-site migration.  At the bench-scale, the deposition of manganese oxides, a permanganate reaction byproduct, has been shown to cause pore plugging and the formation of a manganese oxide layer adjacent to the non-aqueous phase liquid creosote which reduces post-treatment mass transfer and hence mass loading from the source.  The objective of this study was to investigate the potential of partial permanganate treatment to reduce the ability of a coal tar creosote source zone to generate a multi-component plume at the pilot-scale over both the short-term (weeks to months) and the long-term (years) at a site where there is >10 years of comprehensive synoptic plume baseline data available.  A series of preliminary bench-scale experiments were conducted to support this pilot-scale investigation. The results from the bench-scale experiments indicated that if sufficient mass removal of the reactive compounds is achieved then the effective solubility, aqueous concentration and rate of mass removal of the more abundant non-reactive coal tar creosote compounds such as biphenyl and dibenzofuran can be increased.  Manganese oxide formation and deposition caused an order-of-magnitude decrease in hydraulic conductivity. Approximately 125 kg of permanganate were delivered into the pilot-scale source zone over 35 days, and based on mass balance estimates <10% of the initial reactive coal tar creosote mass in the source zone was oxidized.  Mass discharge estimated at a down-gradient fence line indicated >35% reduction for all monitored compounds except for biphenyl, dibenzofuran and fluoranthene 150 days after treatment, which is consistent with the bench-scale experimental results.  Pre- and post-treatment soil core data indicated a highly variable and random spatial distribution of mass within the source zone and provided no insight into the mass removed of any of the monitored species. The down-gradient plume was monitored approximately 1, 2 and 4 years following treatment.  The data collected at 1 and 2 years post-treatment showed a decrease in mass discharge (10 to 60%) and/or total plume mass (0 to 55%); however, by 4 years post-treatment there was a rebound in both mass discharge and total plume mass for all monitored compounds to pre-treatment values or higher. The variability of the data collected was too large to resolve subtle changes in plume morphology, particularly near the source zone, that would provide insight into the impact of the formation and deposition of manganese oxides that occurred during treatment on mass transfer and/or flow by-passing. Overall, the results from this pilot-scale investigation indicate that there was a significant but short-term (months) reduction of mass emanating from the source zone as a result of permanganate treatment but there was no long-term (years) impact on the ability of this coal tar creosote source zone to generate a multi-component plume.

Xu, X., N.R. Thomson, 2008. Estimation of the maximum consumption of permanganate by aquifer solids using a modified chemical oxygen demand test, ASCE J. Envir. Eng., 134(5), 353-360.

Knowledge of the consumption of permanganate by naturally occurring reduced species associated with aquifer materials is required for site screening and design purposes to support permanganate in situ chemical oxidation (ISCO) applications. It has been established that this consumption is not a singled-valued quantity but rather is kinetically controlled. Current methods to determine this permanganate natural oxidant demand (NOD) involve the use of well-mixed batch tests which are time consuming and subject to test variables (e.g., concentration, mass of oxidant to solid ratio, reaction duration, and mixing conditions) that significantly affect the results. In this paper we propose a modified chemical oxygen demand (COD) test using permanganate which can be used to determine the maximum permanganate NOD of an aquifer material. As an initial point of comparison, we tested aquifer materials collected from eight (8) potential ISCO sites using this modified or permanganate COD method, the traditional dichromate COD method, and a method based on well-mixed batch reactors. The results from this comparison indicated that there was no statistically significant difference (α = 5%) between the results of the permanganate COD test and the maximum NOD from the well-mixed batch reactors, while on average the dichromate COD test overestimated the maximum NOD by 100%. The permanganate COD test results were highly correlated to the batch-test maximum NOD data (r = 0.996), and to the total organic carbon and amorphous Fe content of the aquifer materials (r = 0.91). A limited sensitivity investigation of this proposed permanganate COD test revealed that the suspected formation of manganese oxides, a reaction by-product, may lead to increased experimental variability. However, in spite of this concern we recommend that this proposed permanganate COD method is a quick and economical approach for estimating the maximum permanganate NOD for aquifer materials to support permanganate ISCO site screening and initial design purposes.

Thomson, N.R., G. Roos, R.D. Wilson, S.F. Thornton, B.A. Tolson, 2008. The dipole flow and reactive tracer test for aquifer parameter estimation.  Presented at the AGU Fall Meeting, San Francisco, CA, Dec 15-19.

Effective and efficient contaminated site remediation requires site-specific knowledge of physical, chemical and biological properties of the aquifer (e.g., hydraulic conductivity, porosity, ion exchange capacity, redox capacity, and biodegradation potential).  Aquifer property measurement techniques for groundwater transport and reactions are too costly or not-representative of in situ conditions and therefore there is an over-reliance on literature values or model assumptions.  This results in overly uncertain predictions of in situ performance and therefore unnecessarily cautious risk assessment and costly remediation strategies.  Therefore, cost-effective site investigative tools that have the capability of producing high quality characterization data are required.  The dipole flow and reactive tracer test (DFRTT) is proposed as an alternative to current parameter estimation methods which may be biased or non-existent.  The DFRTT circulates groundwater between isolated injection (source) and extraction (sink) chambers within a single well.  Once steady-state flow has been reached, conservative and reactive tracers are added to the injected solution and the concentration of the tracers and their reaction products can be monitored in the extracted solution.  These tracer breakthrough curves are analyzed by the DFRTT interpretation model to obtain parameter estimates.  A dipole probe prototype has been constructed at the University of Waterloo and more than 50 field tests have been conducted in the unconfined sand aquifer at CFB Borden near Alliston, ON.  Four characteristic type or response curves were observed.  The DFRTT showed good repeatability between tests and captures both the response of the disturbed zone and formation.  The DFRTT response profiles were determined to be scalable to some of the key system design parameters.  Estimates of the hydraulic conductivity of the aquifer were within literature values, but variability at the sensing scale of the dipole tool was observed.  In addition to an overview of this project, a review of the 2007-2008 field results along with model interpretation efforts are discussed.

Sra, K., N.R. Thomson, J.F. Barker, 2008.  In situ chemical oxidation of gasoline compounds using persulfate, Proceedings of the Petroleum Hydrocarbons and Organic Chemicals in Ground Water Conference, Houston, TX, Nov. 3-5.

Destruction of gasoline compounds and fractions – benzene, toluene, ethylbenzene, xylenes (BTEX), trimethylbenzenes (TMBs) and naphthalene, gasoline fractions (F1 and F2) and total petroleum hydrocarbons (TPH) – by activated and unactivated persulfate was studied at the bench-scale. Unactivated batch reactor systems employed persulfate at 1 or 20 g/L and activated systems employed persulfate at 20 g/L and H2O2 as activator at two experimental conditions - 0.1 or 1.0 mol H2O2/mol S2O82-. All treatments and controls contained API standard gasoline at ~25 mg/L and were run in triplicate. Sampling was conducted over a ~28 day period. Controls showed insignificant degradation for all gasoline compounds and fractions examined while unactivated persulfate at 1 g/L showed small (<10%) decrease in concentration of gasoline compounds over the reaction period. Unactivated persulfate at 20 g/L demonstrated a large decrease in concentration of BTEX (>99%), TMBs (>94%) and naphthalene (>71%). Oxidation of F1 (>94%) seemed to be more impacted than F2 (>80%) while >93% TPH was oxidized. Use of peroxide as activator at 0.1 mol H2O2/mol S2O82- improved the conversion of TMBs (>99%) and naphthalene (>85%). Increase in activator strength to 1.0 mol H2O2/mol S2O82- decreased the conversion for xylenes (>86%) and TMBs (>81%) implying that an optimal H2O2 molar ratio with persulfate may be required to maximize treatment of gasoline compounds. With H2O2 alone at the two conditions, the treatment of compounds was higher for molar ratio 1.0 (<27%) than for molar ratio 0.1 (<11%). Overall, while persulfate at 20 g/L alone removed >92% TPH, H2O2 alone at the same molar ratio oxidized only ~17% TPH. Use of persulfate at ~20 g/L by itself or in combination with optimal doses of peroxide seems to be a viable option for remediation of gasoline compounds examined in this study.

O’Connell, M., N.R. Thomson, 2008. Optimizing Solvent Extraction of PCBs in the Field. Proceedings from the 24th Annual International Conference on Soils, Sediments and Water University of Massachusetts, Amherst, MA, Oct 20-23.

Polychlorinated biphenyls (PCBs) are carcinogenic persistent contaminants and although
their manufacturing in North America ceased in the late 1970s, their high heat resistance
made their use widespread over their production lifetime. PCB contaminated soil or sediments on industrial and other sites have historically been dealt with through
excavation followed by off-site disposal or incineration. One potential technology that
has shown some demonstrated success in the southern United States is solvent extraction. Although successful at removing a large quantity of PCBs from soil, this technology can be improved upon by making the extraction more complete, efficient and suitable for in a variety of climates. Research underway at the University of Waterloo will identify the factors controlling PCB extraction with solvents in order to optimize PCB extraction as it is applied on different soil types and in various climates. The collected data will be used to develop a temperature-corrected kinetic model to better represent the extraction process. As past research has shown that weathered PCB in soil is more difficult to remove, contaminated field samples from Southern Ontario, Canada are being used for this work, rather than synthetically prepared samples. Initial screening experiments consisted of mixing a solvent with contaminated soil while varying factors potentially influencing the extraction, such as moisture content, solvent type, and grain size. Additional experiments were conducted to discover if the influence of factors controlling PCB extraction was temperature dependent. This presentation will include a discussion of the experiments underway, factors influencing solvent extraction, and the expected form of the kinetic model.

Sra, K., N.R. Thomson, J.F. Barker, 2008.  In situ chemical oxidation of gasoline compounds using persulfate, REMTECH 2008, Banff, AB, Oct. 15-17.

In situ chemical oxidation (ISCO) using persulfate (Eo = 2.01 V) is a promising remediation technology that can be potentially applied to a wide range of organic contaminants. Gasoline compounds are of particular interest because they extensively impact the soil and groundwater and are highly persistent and toxic. In this investigation, destruction of gasoline compounds and fractions – benzene, toluene, ethylbenzenes, xylenes (BTEX), trimethylbenzenes (TMBs) and naphthalene, gasoline fractions (F1 and F2) and total petroleum hydrocarbons (TPH) – by activated and inactivated persulfate was studied at the bench-scale. Aqueous phase batch reactors (25 mL) for inactivated systems employed persulfate at two concentration (1 or 20 g/L) and activated systems were conducted at 20 g/L persulfate concentration. In activated systems, utility of hydrogen peroxide as activator was examined at two experimental conditions (molar ratio 0.1 and 1.0 with respect to persulfate). Reactors were set up to determine individual impact of peroxide at the two conditions. All treatments and controls contained gasoline at ~25 mg/L and were run in triplicate. Sampling for gasoline compounds was conducted over a ~28 day reaction period. Controls showed insignificant degradation for all the gasoline compounds and fractions examined while inactivated persulfate at 1 g/L showed little (<10%) decrease in concentration of gasoline compounds over the reaction period. Inactivated persulfate at 20 g/L demonstrated a large decrease in aqueous concentration of BTEX (>99%), TMB (>94%) and naphthalene (>71%). Oxidation of F1 (>94%) seemed to be more impacted than F2 (>80%) while >93% TPH was oxidized. Use of peroxide as activator at molar ratio of 0.1 improved the conversion of TMBs (>99%) and naphthalene (>85%) while maintaining a high conversion for BTEX (>99%) compounds. Increase in activator strength (molar ratio 1.0) decreased the conversion for xylenes (>86%) and TMBs (>81%). This implies that an optimum range of activator conditions may be required to maximize oxidation of gasoline compounds by peroxide-activated persulfate. Use of peroxide alone at the two conditions led to a relatively small change in concentration of these compounds. The decrease in concentration of all the compounds was higher for molar ratio 1.0 (<27%) as compared with molar ratio condition 0.1 (<11%). On a molar basis, persulfate exhibits a higher efficiency than peroxide in treatment of these gasoline compounds. Overall, while persulfate at 20 g/L removed >92% TPH, peroxide at the same molar ratio oxidized only ~17% TPH. Use of persulfate at high dosages by itself or in combination with optimum doses of peroxide as activator seems to be a viable option for remediation of gasoline compounds examined in this study. Persulfate appears to be particularly effective in the oxidation of BTEX compounds and may require activation for a near complete oxidation of TMBs. Highest conversion of naphthalene was observed with peroxide-activated persulfate. Activation of persulfate enhanced the oxidation of F1 compounds whereas F2 compounds appear to have a 60 – 70% oxidizable fraction which was removed in all treatments employing persulfate or peroxide.

Roos, G., N.R. Thomson, 2008. The dipole flow and reactive tracer test for aquifer parameter estimation.  Proceeding from the 61st Canadian Geotechnical Conference and the 9th Joint CGS/IAH-CNC Groundwater Conference, Edmonton, AB, Sept. 21-24.

Effective and efficient groundwater remediation requires site-specific knowledge of physical, chemical and biological aquifer properties.  The dipole flow and reactive tracer test (DFRTT) is single well test proposed as a cost-effective alternative to current aquifer parameter estimation methods.  A steady-state dipole flow field is created by circulating groundwater between chambers isolated by the dipole tool.  A tracer is released into the injection chamber and the breakthrough curve at the extraction chamber is interpreted with the DFRTT specific model.  This paper describes the construction of a prototype dipole tool and results from ongoing field trials.

Sra, K., N.R. Thomson, J.F. Barker, 2008.  Persistence of peroxide-activated persulfate, Remediation of Chlorinated and Recalcitrant Compounds: Sixth International Conference, Monterey, CA, May 19-22.

Activated persulfate exhibits a wide range of contaminant reactivity and appears to be relatively stable in the subsurface; however, issues related to the development of a cost-effective activation scheme and estimation of in situ persulfate decomposition rates remain unresolved.  Contemporary persulfate activation strategies (e.g. ferrous iron, heat, UV, pH) either lead to very low persulfate persistence or may require a high degree of engineering for implementation at full scale. Use of peroxide to activate persulfate is the newest proposed scheme and hence little is known on its impact on the persistence of persulfate at the bench or pilot-scale. To quantify the impact of peroxide activation on the fate and transport of persulfate in the presence of aquifer materials, a series of bench and pilot-scale investigations were conducted.  The bench-scale experiments employed well characterized aquifer materials from four sites across North America.  Batch tests were employed at various activator (peroxide) to oxidant (persulfate) mass ratios (0, 0.01, 0.1, 1 and 10) and at two persulfate concentrations (1 and 20 g/L). For a given aquifer material, the total oxidation strength decreased rapidly at early times followed by a relatively slower rate of depletion for all mass ratios explored suggesting that the early depletion was controlled by peroxide.  Results from these bench-scale experiments are expected to provide data that will be used to establish optimum peroxide to persulfate mass ratios and to identify controlling factors for the range of aquifer materials studied. To further investigate persulfate stability under conditions more representative of the subsurface environment, pilot-scale experiments consisting of push-pull studies were conducted at the aquifer research facility at CFB, Borden.  Push-pull tests involved the injection of persulfate at (1 g/L), peroxide (at mass ratio 0 and 1 with respect to persulfate) and a conservative tracer into a hydraulically isolated portion of the sandy aquifer.  Groundwater samples were extracted over time and analyzed for total oxidant strength and conservative tracer concentration to yield total oxidant degradation rates. This investigation will yield persistence data at multiple scales that can be used to develop persulfate degradation rate scale-up factors (with and without activation) so that robust and cost-effective bench-scale experiments can be used to estimate in situ persulfate degradation rates.

Sra, K., N.R. Thomson, J.F. Barker, 2007. Fate of persulfate in uncontaminated aquifer materials. Presented at Groundwater Quality 2007, Fremantle, Australia, Dec. 2–7, 2007.

A successful and cost-effective persulfate in situ chemical oxidation system depends not only on the delivery scheme employed but also on a comprehensive understanding of the interaction of persulfate with the target contaminant(s) and aquifer materials.  To improve our understanding of the in situ persistence of persulfate and to identify potential controlling factors required for oxidant loading estimates, a series of bench and pilot-scale experiments were performed to quantify the interaction of persulfate with aquifer materials.  Well-characterized materials from seven sites across North America were used in a series of bench-scale studies to estimate persulfate decomposition kinetic parameters and to quantify the impact on aquifer material properties.  The batch experiments were conducted in an experimental system containing 100 g of solids and 100 mL of persulfate solution at 1 or 20 g/L.  Column experiments, more representative of in situ conditions with respect to oxidant to aquifer solids mass ratio, were performed in a stop-flow mode using a 1 g/L persulfate solution.  Results from the bench-scale experiments indicated that the decomposition of persulfate followed a first-order rate law for all aquifer materials we investigated.  An order of magnitude decrease in decomposition rate coefficients were observed for systems that used a persulfate concentration of 20 g/L as compared to those that used 1 g/L.  As expected, the column experiments yielded higher reaction rate coefficients than batch tests for the same persulfate concentration due to the higher oxidant to solids mass ratio and increased contact area. At the pilot-scale, a series of push-pull tests and a three-dimensional transport study were performed to quantify persulfate stability.  The push-pull tests involved the injection of persulfate (1 or 20 g/L) and a conservative tracer into a hydraulically isolated portion of the sandy aquifer at CFB Borden followed by the slow extraction of groundwater for ~15 days. Analyses of the conservative tracer and persulfate concentrations yielded in situ persulfate decomposition rates which were ~50 times higher than those estimated from the bench-scale tests. The three-dimensional transport study involved a near continuous point injection of persulfate into the CFB Borden aquifer followed by monitoring its breakthrough at multilevel sampling fence lines located 3 and 9 m down-gradient. Estimation of the reduced persulfate mass discharge at each fence line was used to quantify a persulfate decomposition rate at the pilot-scale. Overall, the estimated rate coefficients indicate that persulfate is persistent and hence a suitable oxidant for the range of aquifer materials explored in this study. The findings from this multi-scale investigation were used to develop persulfate decomposition scale-up factors so that robust and cost-effective bench-scale techniques can be used to estimate in situ persulfate degradation rates.

Thomson, N.R., E. D. Hood, G. J. Farquhar, 2007. Permanganate treatment of an emplaced DNAPL source. Ground Water Monitoring and Remediation, Fall 2007.

In situ chemical oxidation (ISCO) using permanganate is one of the few promising technologies that have recently appeared with the capability of aggressively removing mass from non-aqueous phase liquid (NAPL) source zones.  While NAPL mass in regions of the treatment zone where delivery is dominated by advection can be removed rather quickly, the rate of mass removal from stagnant zones is diffusion controlled.  This gives rise to partial mass removal and a concomitant reduction in the NAPL mass, down-gradient groundwater concentrations, and the dissolution rate associated with the source zone.  Therefore monitoring the performance of a permanganate ISCO treatment system is important to maintain the desired efficiency and to establish a treatment end-point. In this paper we illustrate the use of various monitoring approaches to assess the performance of a pilot-scale investigation that involved treatment of a multi-component NAPL residual source zone with permanganate using a groundwater recirculation system for 485 days. On-going treatment performance was assessed using permanganate and chloride concentration data obtained from extraction wells, 98 piezometers, and groundwater profiling.  At the completion of treatment 23 intact soil cores were extracted from the source zone and used to determine the remaining NAPL mass, and manganese deposition.  Based on the data collected, >99% of the initial NAPL mass was removed during treatment; however, remnant NAPL was sufficient to generate a small but measurable dissolved phase trichloroethene (TCE) and perchloroethene (PCE) plume.  As a result of treatment the ambient-gradient discharge rates were reduced by 99% for TCE and 89% for PCE relative to baseline conditions. The lack of complete source zone oxidation was presumed to be the result of dissolution fingers which channelled the permanganate solution through the source zone preventing direct contact with the NAPL and giving rise to diffusion-limited mass removal.

Nelson, L., C. Doughty, T. Li., J. Barker, M. Ioannidis, N.R. Thomson, J. Chatzis, Field trail of residual LNAPL recovery using CO2 supersaturated water injection, Proceeding from Remediation Technologies Symposium, Banff, AB, October 24 - 26, 2007.

Supersaturated water injection (SWI) has been shown to both volatilize and mobilize residual non-aqueous phase liquid (NAPL) ganglia in laboratory experiments. This technology uses Gas inFusion™ technology to effi ciently dissolve gases into liquids at elevated pressures. During SWI, pressurized water containing high concentrations of CO2 is injected into the subsurface below the zone of contamination. When the injected water enters the aquifer, the pressure drops and CO2 bubbles nucleate. These bubbles then migrate upwards through the contaminated zone towards the water table. As they move they come into contact with residual NAPL ganglia and they either volatilize or mobilize the NAPL. In either case the bubbles continue to rise until they reach the water table at which point they are removed from the subsurface by a multi-phase extraction system. In this project, the ability of SWI to recover NAPLs was studied at the fi eld scale as part of an ongoing program to evaluate the applicability of this technology to groundwater remediation. To accomplish this, a know amount of NAPL was emplaced below the water table at residual concentrations to represent a residual source of gasoline. The NAPL contained 80 litres of pentane, 80 litres of hexane and 40 litres of Soltrol 130. The source was created in hydraulically isolated cell in an unconfi ned sand aquifer. After the source was emplaced, SWI was used in an attempt to remove the contaminant mass. In order to quantify the mass of each of the contaminants removed from the cell, vapour and water samples were taken frequently and the rate at which water and vapour were extracted from the cell was monitored. The goal of this project was to determine if SWI was capable of removing residual NAPL at a field site. It was successful in removing volatile NAPL but not non-volatile NAPL. Although it is possible that non-volatile NAPL was mobilized to the water table but not removed from the cell. 64% of the volatile compounds were removed but contaminant mass was still being removed when the system was shut down. These results indicate that continued development of the technology for groundwater remediation is warranted.

Bekeris, L., D.L. Rudolph, N.R. Thomson, 2007. Field-scale evaluation of enhanced agricultural management practices using a novel unsaturated zone nitrate mass flux approach.  In Proceeding of  the 60th Canadian Geotechnical Conference and 8th Joint CGS/IAH-CNC Specialty Groundwater Conference, Ottawa, ON, Oct. 21-24, 2007.

A field technique for measuring unsaturated zone nitrate mass flux was developed and applied in an agricultural setting to assess the effect of a reduction in nitrogen application. Nitrate mass flux was calculated as the product of groundwater recharge rate and porewater nitrate concentration. Results of this novel approach indicated that several factors may limit its initial application in quantifying mass flux changes. Strategies for improving the method’s utility are discussed. Results illustrated, however, that significant reductions in stored nitrate mass could be observed at several locations following the reduction in nutrient application.

Yasuda, N., N.R. Thomson, J.F. Barker, 2007. Performance evaluation of a tailings pond seepage collection system. In Proceeding of the 60th Canadian Geotechnical Conference and 8th Joint CGS/IAH-CNC Specialty Groundwater Conference, Ottawa, ON, Oct. 21-24, 2007.

The evaluation of a tailings pond seepage collection system was performed at an active oil sands mining site.  A network of piezometers and drive points were installed in a 1 km2 area to facilitate hydraulic measurements and water sampling to characterize the surface water and groundwater flow system.  Chemical tracers suggest migration of process affected water in a shallow, permeable sand deposit beyond an inner seepage collection ditch, but elevated hydraulic heads beyond the outer ditch have prevented further migration.  Under the present hydraulic conditions the seepage collection system is currently working to effectively contain process affected water.

Rahman, R., E.O. Frind, D.L. Rudolph, N.R. Thomson, J.W. Molson, 2007. Contaminant risk assessment for estimating the cost of water well remediation or replacement.  In Proceeding of  the 60th Canadian Geotechnical Conference and 8th Joint CGS/IAH-CNC Specialty Groundwater Conference, Ottawa, ON, Oct. 21-24, 2007.

The risk of well contamination due to potential sources within a capture zone can be assessed using a quantitative approach. The approach is based on the concept of well vulnerability, which includes all key characteristics and processes such as the nature of the source, the transport and fate of the contaminants along its path from the source to the receptor, and the interaction of the well itself with the flow system. The results provide the exposure value and the time frame within which the well will become contaminated, and can be expressed in terms of the investment that must be made to cover the future cost of well remediation.

Jones, L., N.R. Thomson, X. Xu, 2007. A permanganate natural oxidant demand kinetic model.  Presented at the 23nd Annual Conference on Soils, Sediments, and Water, Univ. of Mass., Amherst, MA, Oct. 15-18, 2007.

The presence of naturally occurring reduced species associated with aquifer materials exerts a significant permanganate demand thereby reducing the mass of oxidant available for the destruction of the contaminant(s) of concern as well as reducing the oxidation rate.  Recent laboratory efforts indicate that this demand is not a single-valued quantity, but is kinetically controlled and depends on the parameters of the test system and type of reduced aquifer material species present.  A comprehensive model that captures the kinetic behaviour of permanganate consumption by aquifer materials was formulated by using data collected from well-mixed batch reactor and column systems.  The batch experiments were based on the theoretical derivation of the stoichiometric reaction of permanganate with bulk aquifer material reductive components, and consisted of excess permanganate mass experiments and excess aquifer material mass experiments.  A typical experimental column trial consisted of flushing an aquifer-material packed column with the permanganate source solution until sufficient permanganate breakthrough was observed.  Aquifer material from several representative sites across North America was used.  We assumed that the dichromate chemical oxygen test results could serve as a surrogate for the overall aquifer material reduction capacity.  The developed kinetic model consists of three reactive components associated with the aquifer material: a fast component, an intermediate component, and a slow component.  The fast and intermediate components were observed in the batch experiments, while the slow component was observed in the column breakthrough curves.  Evidence of passivation was apparent in the data and confirmed by manganese oxide coating on grains.  This presentation will include a discussion of the underlying processes, and a description of the experimental data and aspects of the developed kinetic model.  In addition, the impact of permanganate consumption kinetics on source zone treatment will be demonstrated.

Xu, X., N.R. Thomson, 2007. Stability of hydrogen peroxide in the presence of aquifer solids using a green chelating reagent, Chemosphere, 69, 755-762. doi:10.1016/j.chemosphere.2007.05.008.

Hydrogen peroxide is a widely used in situ chemical oxidation reagent which relies on catalysts to generate the suite of reactive species that are required to aggressively remediate contaminated soils and groundwater.  In the subsurface environment these catalysts are usually transition metals that are added to the injected solution, or are naturally occurring.  Chelating agents are widely used to maintain an adequate dissolved transition metal concentration in near-neutral pH conditions; however, they can also be used to improve the persistence of H2O2 in situations when the aquifer solids have sufficient transition metal content.  Ethylenediamine tetraacetic acid (EDTA) and nitrilotriacetic acid (NTA) have been considered to be the most effective chelants and therefore are the most widely used.  While previous research efforts have focused on the chelating agent efficiency, the long-term fate of these chelants in the natural subsurface environment is a concern since both EDTA and NTA are non-readily biodegradable.  The focus of this investigation was to evaluate the potential of using the environmentally friendly or green chelating agent ethylenediaminedisuccinate (EDDS) as an alternative to EDTA or NTA to suppress the catalytic activity of naturally-occurring transition metals.  A series of batch reactor and column experiments were performed using five different aquifer materials and the results demonstrate that EDDS has a comparative chelating efficiency to that of EDTA.  The addition of EDDS was able to reduce the H2O2 decomposition rates in the presence of the aquifer materials used in this investigation by 24 to 97% in well-mixed batch systems, and by 20 and 38% in the column trials where H2O2 was detected in the effluent.

Cai, Z., C. Merly, N.R. Thomson, R.D. Wilson, D. Lerner, 2007. Channel flow and reactive transport in a partly iron-filled fracture: Experimental and model results, J. Contam. Hydrol., 93, 284–303, doi:10.1016/j.jconhyd.2007.04.001.

Technical developments have now made it possible to emplace granular zero-valent iron (Fe0) in fractured media to create a Fe0 fracture reactive barrier (Fe0 FRB) for the treatment of contaminated groundwater.  To evaluate this concept, we conducted a laboratory experiment in which trichloroethylene (TCE) contaminated water was flushed through a single uniform fracture created between two sandstone blocks.  This fracture was partly filled with what was intended to be a uniform thickness of iron.  Partial treatment of TCE by iron demonstrated that the concept of a Fe0 FRB is practical, but was less than anticipated for an iron layer of uniform thickness.  When the experiment was disassembled, evidence of discrete channelized flow was noted and attributed to imperfect placement of the iron.  To evaluate the effect of the channel flow, an explicit Channel Model was developed that simplifies this complex flow regime into a conceptualized set of uniform and parallel channels.  The mathematical representation of this conceptualization directly accounts for (i) flow channels and immobile fluid arising from the non-uniform iron placement, (ii) mass transfer from the open fracture to iron and immobile fluid regions, and (iii) degradation in the iron regions.  A favourable comparison between laboratory data and the results from the developed mathematical model suggests that the model is capable of representing reactive transport in fractures with non-uniform iron placement.  In order to apply this Channel Model concept to a Fe0 FRB system, a simplified, or implicit, Lumped Channel Model was developed where the physical and chemical processes in the iron layer and immobile fluid regions are captured by a first-order lumped rate parameter.  The performance of this Lumped Channel Model was compared to laboratory data, and validated against the Channel Model.  The advantages of the Lumped Channel Model are that is has only one parameter that can be used directly in readily available numerical simulators.

Cai, Z., N.R. Thomson, R.D. Wilson, S. Oswald, 2006. A lumped parameter approach to model the treatment of organic contaminants by a granular iron filled fracture, Adv. Water Resour., 29, 624–638.

Zero-valent iron permeable reactive barriers (Fe0-PRBs) have been widely used in unconsolidated media to treat certain dissolved phase organic contaminants, but little attention has been given to their application in fractured porous media.  In principle, it is possible to create a Fe0-PRB in a fractured porous medium by injecting an iron slurry mixture into the fracture network.  This emplacement method likely results in a complicated system of incompletely iron filled fractures.  To aid in the design and performance assessment of such complex systems, representative models must be used that capture the essence of the controlling processes; however, existing models cannot directly account for the complex processes dominating treatment in a partly iron-filled fracture at the required spatial scale.  As a first step to address this need, we have developed a modelling approach for an idealized single partly iron filled fracture wherein the physical and chemical processes are represented by a first-order lumped rate parameter.  The performance of the developed lumped rate parameter model was examined over a range of conditions by comparing simulation results to those produced by a more comprehensive analytical solution and a numerical model.  While some deviations were observed, the lumped parameter model was shown to be valid for a range of iron grain sizes, iron layer thicknesses, open fracture apertures, flow velocities, and reaction rate coefficients.  We also demonstrated that the developed lumped parameter approach can represent situations where the system is initially contaminated, and can be used to optimize the thickness of the iron layer.  The advantage of this first-order lumped rate parameter model is that it can be used directly in existing discrete fracture models without modifications to their computational framework, and hence will make it possible to approximate the field-scale treatment performance of Fe0-PRBs in fractured porous media.

Sra, K.S., N.R. Thomson, J.F. Barker, 2006. Persulfate decomposition kinetics in the presence of aquifer materials.  In the Proc from the 22nd Annual Conference on Soils, Sediments, and Water, University of Massachusetts, Amherst, MA, October 16-19, 2006.

Persulfate is an emerging oxidant for in situ chemical oxidation (ISCO) applications with a high oxidation potential on activation (E = 2.6 V).  The design of an oxidant remedial system involves a comprehensive understanding of a number of underlying physical and chemical processes.  One of these processes, which impacts oxidant efficiency, is the stability of the oxidant in the presence of natural aquifer materials.  To improve our understanding and develop predictive relationships, a series of batch and column experiments were designed to analyse and quantify the impact of reductants including total organic carbon, iron, and manganese on the decomposition of persulfate.  Well-characterized aquifer materials collected from nine sites across North America were used in this investigation.  The batch experiments were conducted to primarily observe and derive decomposition kinetic parameters under a range of initial persulfate concentrations and oxidant mass to solids loading ratios.  The column experiments, which are more representative of in situ conditions, consisted of 40 cm long columns packed with aquifer material and flushed with various persulfate concentrations at several flow rates.  Observed high decomposition rates indicate poor oxidant efficiency and implies that the site may be unsuitable for persulfate ISCO treatment, while a very low decomposition rate indicates a site where an external source of activation is required.  This presentation will provide an overview of the experimental design and results, and discuss the utility of the various predictive relationships developed between the decomposition kinetics and various aquifer material properties and experimental variables.  

Jones, L., X. Xu, N. R. Thomson, R. Waldemer, P.G. Tratnyek, 2006.  The impact of permanganate NOD kinetics on treatment efficiency. In Proceedings from the Fifth International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, California, May 22-25, 2006.

Permanganate natural oxidant demand (NOD) is the unproductive consumption of permanganate by the reduced species associated with aquifer solids and requires quantification for site screening and system design.  Recent laboratory efforts indicate that this demand is not a single-valued quantity, but is kinetically controlled and depends on the test system parameters and type of reduced aquifer material species present.  As a first step to understand how NOD kinetics may affect the oxidation of residual non-aqueous phase liquids, a series of one-dimensional mathematical simulations were performed in which both experimentally observed NOD and organic oxidation kinetics were employed.  The NOD kinetic expression was based on data collected from a series of laboratory experiments involving well-mixed batch reactor and column systems, while the organic oxidation reaction rate coefficients were obtained either from recent laboratory experiments or relevant literature.  The results indicate that NOD kinetics appear to play an insignificant role in the overall treatment efficiency in the scenario we explored; however, these results are preliminary and additional research is ongoing to elucidate the role of permanganate NOD kinetics in ISCO applications.

Xu, X., N. R. Thomson, 2006. Oxidant fate in the subsurface environment: From batch to column systems. In Proceedings from the Fifth International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, California, May 22-25, 2006.

While use of in situ chemical oxidation (ISCO) as a remediation strategy has gained widespread acceptance due to its advantages over other conventional treatment technologies, our knowledge relating to the behaviour of common oxidants (e.g., permanganate, hydrogen peroxide, and persulfate) in the subsurface has not developed as quickly.  Delivery of the chosen oxidant to the target treatment area must occur in situ and hence the stability of the injected oxidant in the presence of naturally occurring reactive reductants (e.g., natural organic matter and reduced inorganic species) and catalysts associated with aquifer solids is a controlling factor for an ISCO system to be successful.  For permanganate the natural oxidant demand (NOD) is usually estimated from a bench-scale test using uncontaminated aquifer material samples.  The reported NOD values represent the consumption of the oxidant by reactions unrelated to the compound of concern, and thus capture the impact of the naturally occurring reactive reductants.  These values are used as the basis for oxidant dosing and in some cases for a “go/no-go” decision.  For catalyzed hydrogen peroxide a NOD value cannot be determined but rather an enhanced decomposition rate can be estimated.  A considerable effort has been expended by numerous research groups and consultants to provide a bench-scale estimate of these important oxidant/aquifer material interactions.  In most cases well-mixed batch systems characteristic of a soil slurry mixture are used and are not representative of subsurface conditions in which these heterogeneous reactions occur.  An important question is how results from well-mixed batch systems represent results from flow-through column systems which are considered more physically representative of in situ conditions. To provide insight into potential scaling relationships a series of carefully-designed batch and column experiments were performed to examine the fate of permanganate and hydrogen peroxide in the presence of nine uncontaminated aquifer materials collected from sites across North America.  Batch tests for permanganate consumption consisted of 40 mL reactors filled with a mixture of permanganate solution and aquifer materials, while batch tests for hydrogen peroxide decomposition were performed in 100 mL ample reactors with an initial pH established ~3.0.  For the column tests, permanganate solutions were flushed through aquifer materials packed in transparent Plexiglas columns (40 cm long and 5 cm ID) with multiple sampling ports, and hydrogen peroxide solutions were flushed through aquifer materials packed in HDPVC columns (10 cm long and 3 cm ID). Results from these experiments indicate that hydrogen peroxide decomposition rates extracted from batch tests significantly underestimate observed column decomposition rates, and batch test permanganate consumption overestimates column derived permanganate consumption values.  Based on these findings, we have attempted to develop simple relationships to use the information obtained from batch tests to capture the column test observations.  These relationships will allow us to extend the results obtained from relatively fast and economical batch tests to more physically representative conditions.

Murphy, J. R., N. R. Thomson, 1993. Two-phase flow in a variable aperture fracture, Water Resour. Res., 29(10), 3453-3476.

In this paper a dynamic two-dimensional two-phase flow model for a single variable aperture fracture is developed. The model is based on a finite volume implementation of the cubic law and the conservation of mass for each liquid. The two-phase fracture flow system is represented by incompressible parallel plate flow within two-dimensional subregions of constant aperture. The fluid phase distribution is represented by an explicit definition of the phase presence at each location within the domain. To achieve this definition, a phase distribution is assigned to each fracture subregion. Knowledge of the phase distribution allows calculation of interface capillary pressure based on the fracture aperture. One-dimensional analytic solutions for two-phase flow are developed and used to verify the model's behavior in one dimension. The model is verified against the Sandia Waste-Isolation Flow and Transport III model for the case of two-dimensional single-phase flow. Two-dimensional two-phase flow verification is performed qualitatively because no suitable analytic or physical model is currently available. Two-dimensional flow phenomena are investigated for variable aperture fractures generated using geostatistical methods. Results from these simulations illustrate the flow processes of phase isolation, pinching off of nonwetting phase globules, nonwetting phase refusal at the edges of tight regions, and downslope migration of a fluid countercurrent to flow of a less dense fluid.