Magdalena M. Krol

Characterization of nZVI Mobility in a Field Scale Test

Nanoscale zerovalent iron (nZVI) particles were injected into a contaminated sandy subsurface area in Sarnia, Ontario. The nZVI was synthesized on site, creating a slurry of 1 g/L nanoparticles using the chemical precipitation method with sodium borohydride (NaBH4) as the reductant in the presence of 0.8% wt. sodium carboxymethylcellulose (CMC) polymer to form a stable suspension. Individual nZVI particles formed during synthesis had a transmission electron microscopy (TEM) quantified particle size of 86.0 nm and dynamic light scattering (DLS) quantified hydrodynamic diameter for the CMC and nZVI of 624.8 nm. The nZVI was delivered to the subsurface via gravity injection. Peak normalized total Fe breakthrough of 71% was observed 1m from the injection well and remained above 50% for the 24 h injection period. Samples collected from a monitoring well 1 m from the injection contained nanoparticles with TEM-measured particle diameter of 80.2 nm and hydrodynamic diameter of 562.9 nm. No morphological changes were discernible between the injected nanoparticles and nanoparticles recovered from the monitoring well. Energy dispersive X-ray spectroscopy (EDS) was used to confirm the elemental composition of the iron nanoparticles sampled from the downstream monitoring well, verifying the successful transport of nZVI particles. This study suggests that CMC stabilized nZVI can be transported at least 1 m to the contaminated source zone at significant Fe(0) concentrations for reaction with target contaminants.

A Field-Validated Model for In Situ Transport of Polymer-Stabilized nZVI and Implications for Subsurface Injection

Nanoscale zerovalent iron (nZVI) particles have significant potential to remediate contaminated source zones. However, the transport of these particles through porous media is not well understood, especially at the field scale. This paper describes the simulation of a field injection of carboxylmethyl cellulose (CMC) stabilized nZVI using a 3D compositional simulator, modified to include colloidal filtration theory (CFT). The model includes composition dependent viscosity and spatially and temporally variable velocity, appropriate for the simulation of push-pull tests (PPTs) with CMC stabilized nZVI. Using only attachment efficiency as a fitting parameter, model results were in good agreement with field observations when spatially variable viscosity effects on collision efficiency were included in the transport modeling. This implies that CFT-modified transport equations can be used to simulate stabilized nZVI field transport. Model results show that an increase in solution viscosity, resulting from injection of CMC stabilized nZVI suspension, affects nZVI mobility by decreasing attachment as well as changing the hydraulics of the system. This effect is especially noticeable with intermittent pumping during PPTs. Results from this study suggest that careful consideration of nZVI suspension formulation is important for optimal delivery of nZVI which can be facilitated with the use of a compositional simulator.

Diffusion of TCE Through Soil-Bentonite Slurry Walls

Diffusion experiments performed using both a dissolved solution of trichloroethylene (TCE) and a pool of free phase TCE adjacent to a simulated soil-bentonite (SB) wall are described. These tests examine a multi-layer system that includes both contaminated sand and a SB barrier. Results obtained from experiments with dissolved TCE as the primary source are shown to be consistent with those obtained with free-phase TCE as the source of contaminant. Diffusion and sorption coefficients of a soil-bentonite slurry wall are reported to be 3.5 × 10−10 m 2 /s and 0 cm 3 /g, respectively. These diffusion and sorption coefficients were used to evaluate the effectiveness of a hypothetical SB slurry wall located adjacent to a TCE spill.

nZVI injection into variably saturated soils: Field and modeling study

Nano-scale zero valent iron (nZVI) has been used at a number of contaminated sites over the last decade. At most of these sites, significant decreases in contaminant concentrations have resulted from the application of nZVI. However, limited work has been completed investigating nZVI field-scale mobility. In this study, a field test was combined with numerical modeling to examine nZVI reactivity along with transport properties in variably saturated soils. The field test consisted of 142L of carboxymethyle cellulose (CMC) stabilized monometallic nZVI synthesized onsite and injected into a variably saturated zone. Periodic groundwater samples were collected from the injection well, as well as, from two monitoring wells to analyze for chlorinated solvents and other geochemistry indicators. This study showed that CMC stabilized monometallic nZVI was able to decrease tricholorethene (TCE) concentrations in groundwater by more than 99% from the historical TCE concentrations. A three dimensional, three phase, finite difference numerical simulator, (CompSim) was used to further investigate nZVI and polymer transport at the variably saturated site. The model was able to accurately predict the field observed head data without parameter fitting. In addition, the numerical simulator estimated the mass of nZVI delivered to the saturated and unsaturated zones and distinguished the nZVI phase (i.e. aqueous or attached). The simulation results showed that the injected slurry migrated radially outward from the injection well, and therefore nZVI transport was governed by injection velocity and viscosity of the injected solution. A suite of sensitivity analyses was performed to investigate the impact of different injection scenarios (e.g. different volume and injection rate) on nZVI migration. Simulation results showed that injection of a higher nZVI volume delivered more iron particles at a given distance; however, the travel distance was not proportional to the increase in volume. Moreover, simulation results showed that using a 1D transport equation to simulate nZVI migration in the subsurface may overestimate the travel distance. This is because the 1D transport equation assumes a constant velocity while pore water velocity radially decreases from the well during injection. This study suggests that on-site synthesized nZVI particles are mobile in the subsurface and that a numerical simulator can be a valuable tool for optimal design of nZVI field applications.

An analysis of a mixed convection associated with thermal heating in contaminated porous media

The occurrence of subsurface buoyant flow during thermal remediation was investigated using a two dimensional electro-thermal model (ETM). The model incorporated electrical current flow associated with electrical resistance heating, energy and mass transport, and density dependent water flow. The model was used to examine the effects of heating on sixteen subsurface scenarios with different applied groundwater fluxes and soil permeabilities. The results were analyzed in terms of the ratio of Rayleigh to thermal Peclet numbers (the buoyancy ratio). It was found that when the buoyancy number was greater than unity and the soil permeability greater than 10(-12) m(2), buoyant flow and contaminant transport were significant. The effects of low permeability layers and electrode placement on heat and mass transport were also investigated. Heating under a clay layer led to flow stagnation zones resulting in the accumulation of contaminant mass and transport into the low permeability layer. The results of this study can be used to develop dimensionless number-based guidelines for site management during subsurface thermal activities.

Multi-dimensional transport modelling of corrosive agents through a bentonite buffer in a Canadian deep geological repository

The use of a deep geological repository (DGR) for the long-term disposal of used nuclear fuel is an approach currently being investigated by several agencies worldwide, including Canadas Nuclear Waste Management Organization (NWMO). Within the DGR, used nuclear fuel will be placed in copper-coated steel containers and surrounded by a bentonite clay buffer. While copper is generally thermodynamically stable, corrosion can occur due to the presence of sulphide under anaerobic conditions. As such, understanding transport of sulphide through the engineered barrier system to the used fuel container is an important consideration in DGR design. In this study, a three-dimensional (3D) model of sulphide transport in a DGR was developed. The numerical model is implemented using COMSOL Multiphysics, a commercial finite element software package. Previous sulphide transport models of the NWMO repository used a simplified one-dimensional system. This work illustrates the importance of 3D modelling to capture non-uniform effects, as results showed locations of maximum sulphide flux are 1.7 times higher than the average flux to the used fuel container.

Transport of polymer stabilized nano-scale zero-valent iron in porous media

This study presents a set of laboratory-scale transport experiments and numerical simulations evaluating carboxymethyl cellulose (CMC) polymer stabilized nano-scale zero-valent iron (nZVI) transport. The experiments, performed in a glass-walled two-dimensional (2D) porous medium system, were conducted to identify the effects of water specific discharge and CMC concentration on nZVI transport and to produce data for model validation. The transport and movement of a tracer lissamine green B® (LGB) dye, CMC, and CMC-nZVI were evaluated through analysis of the breakthrough curves (BTCs) at the outlets, the time-lapsed images of the plume, and retained nZVI in the sandbox. The CMC mass recovery was >95% when injected alone and about 65% when the CMC-nZVI mixture was used. However, the mean residence time of CMC was significantly higher than that of LGB. Of significance for field implementation, viscous fingering was observed in water displacement of previously injected CMC and CMC-nZVI. The mass recovery of nZVI was lower (<50%) than CMC recovery due to attachment onto sand grain surfaces. Consecutive CMC-nZVI injections showed higher nZVI recovery in the second injection, a factor to be considered in field trials with successive CMC-nZVI injections. Transport of LGB, CMC, and nZVI were modeled using a flow and transport model considering LGB and CMC as solutes, and nZVI as a colloid, with variable solution viscosity due to changes in CMC concentrations. The simulation results matched the experimental observations and provided estimates of transport parameters, including attachment efficiency, that can be used to predict CMC stabilized nZVI transport in similar porous media, although the extent of viscous fingering may be underpredicted. The experimental and simulation results indicated that increasing specific discharge had a greater effect on decreasing CMC-nZVI attachment efficiency (corresponding to greater possible travel distances in the field) than increasing CMC concentration.

Transient modelling of sulphide diffusion under conditions typical of a deep geological repository

ABSTRACT The Canadian Nuclear Waste Management Organization (NWMO) has chosen a used fuel container (UFC) design that consists of an inner steel core and outer copper coating that resists corrosion in the anaerobic underground. However, in these anaerobic conditions, microbiologically influenced corrosion (MIC) can contribute to UFC corrosion if sulphides are present in the repository. Therefore, an engineered barrier system (EBS) consisting of bentonite blocks and pellets has been designed to inhibit production of sulphide by microorganisms and impede the movement of sulphides from the host rock to the UFC. Initially, the bentonite within the deep geological repository will contain no sulphides and it will take time for sulphide from the host rock to saturate the repository. To examine the transient effect of sulphide transport and the effectiveness of the EBS, a 3D numerical model was developed capable of simulating the diffusive transport of sulphide within the NWMO repository. The model was developed using COMSOL Multiphysics, a finite element software package. The developed model included the NWMO placement room, as well as a stand-alone UFC package. Conservative assumptions such as constant-concentration boundary condition and transport-limited corrosion were assumed. The time-dependent analysis predicted the inflow of sulphide into the EBS over time, and the results showed that a steady-state condition would be reached in over 2,000 years in the placement room. The model also showed that the highest sulphide flux occurred at the semi-spherical end caps of the UFC and consequently, that corrosion from sulphide diffusion would not be uniform over the container surface. This paper is part of a supplement on the 6th International Workshop on Long-Term Prediction of Corrosion Damage in Nuclear Waste Systems.