Throughout the United States, many areas exceed the level of safe ground ozone (O3) concentration. Non-natural emissions made as result of daily human activities and natural emissions react photochemically to produce ground O3 concentration. Variation in ground O3 concentration is controlled by local and regional emissions, synoptic and mesoscale meteorology, and boundary layer chemistry and dynamics. When the right meteorological variables are present, rural areas can have unhealthy air conditions with high levels of ground O3 concentration similar to that of metropolitan areas. Particular ground O3 concentration episodes were analyzed to summarize what meteorological variables constitute a healthy or hazardous ground O3 day. Hourly ground O3 data for Mammoth Cave National Park from the years, 1998 through 2003 were analyzed. Once analyzed, a combination of meteorological variables is used in a simple linear regression to create the coefficients for empirical predictive model based on 1998-2000 data. The meteorological variables included maximum temperature (Tmax), diurnal temperature range (DTR), solar radiation (SR), and daily precipitation (Pd). The meteorological coefficients were then used with the available meteorological data from 2001 through 2003 to predict ground O3 for 2001 through 2003. Certain meteorological variables such as SR are not easily available in most regions and rural sites in the United States. Therefore, SR was excluded from the regression model to see if rural areas can also forecast ground O3 sufficiently. Root mean square error, d-index, and mean absolute error were used to assess the performance of the predictive model. These measures were calculated to find out if a significant relationship between ground O3 and the meteorological variables is present. For example, the d-index was calculated and ranged from 0.81-0.84 for the best regression model performances. This suggests that the predictive ground O3 from 2001 through 2003 is in agreement with observed ground O3 from 2001 through 2003.

Iron permeable reactive barriers (PRBs) have been installed at sites contaminated with various reducible organic and inorganic chemicals, particularly chlorinated solvents, worldwide. Many geochemical factors can affect the performance of iron PRBs. Chemicals such as nitrate and Cr(VI) may act as competing oxidants when co-exist with chlorinated solvents. Previous studies observed declines in the rate of TCE degradation by granular iron in the presence of nitrate. Passive oxide formation on the iron surface and an increase in corrosion potential of the iron were determined to be the mechanisms for the decline. Cr(VI), being a stronger oxidant than nitrate, may have a similar but greater effect on the iron reactivity. In addition, chromium oxide as well as dissolved CaCO₃, a common groundwater constituent, form secondary precipitates and are likely to further affect the iron reactivity. The primary objective of this study was to determine the effects of Cr(VI) and dissolved CaCO₃ on the iron reactivity towards TCE and Cr(VI) reduction and to provide mechanistic explanations for the observation. In addition, the applicability of a modified reactive transport model (Jeen, 2005) to the system in which chromate and CaCO₃ co-exist with TCE was evaluated. <br /><br /> Column experiments, including measurements of corrosion potential and surface film composition using Raman spectroscopy were conducted. Five column tests were carried out with input solutions consisting of different combinations of TCE (5 mg/L), Cr(VI) (10 mg/L) and dissolved CaCO₃ (300 mg/L) for eight months. <br /><br /> The results from the column receiving only Cr(VI) showed that Cr(VI) was reduced rapidly by the granular iron and was not detected beyond 10 cm from the influent end of the column by the end of the experiment. However, Cr(VI) profiles migrated from the influent end further into the column overtime, suggesting progressive passivation of the iron near the influent end of the column. The gradual increase in corrosion potential (up to 180 mV positive shift) at the port 3 cm from the inlet with the migration of Cr(VI) profiles suggests the formation and accumulation of higher valent iron oxides, such as hematite and goethite, together with Cr(III) products on the iron surface, passivating iron material. Raman spectroscopic measurements confirmed the presence of passive iron oxides at the end of the experiment. For the column receiving Cr(VI) + TCE, the co-existence of TCE did not affect Cr(VI) reduction kinetics. However, the presence of Cr(VI) affected TCE degradation significantly. Two segments in the migration of TCE profiles are identified: the first segment near the influent end of the column, where the iron was still active towards Cr(VI) reduction but inactive towards TCE degradation, and the second segment where Cr(VI) was fully removed and the TCE degradation continued to follow pseudo-first-order kinetics. The migrations in Cr(VI) and TCE profiles suggest that iron was passivated by Fe(III)/Cr(III) products, and Cr(VI), being a stronger oxidant, was reduced much more rapidly than TCE. It is expected that the first segment of TCE profiles would extend gradually with the migration of Cr(VI) profiles over time. <br /><br /> When dissolved CaCO₃ was added to the columns with Cr(VI) and TCE, either as single contaminant, or as co-contaminants, the pH values near the influent end of the columns remained relatively low (~ pH 7), thus, the presence of dissolved CaCO₃ resulted in a stable corrosion potential and faster degradation rates of TCE and Cr(VI). Over time, however, Cr(VI) reduction and iron corrosion produced OH- and shifted the carbonate-bicarbonate equilibrium, resulting in the precipitation of secondary carbonate minerals, as detected by Raman analysis in the three columns containing CaCO₃. The precipitation and accumulation of the secondary minerals on the iron surface gradually decreased iron reactivity, as indicated from the progressive migrations of TCE profiles in the column receiving TCE +CaCO₃ and of second segment TCE of profiles in the column receiving TCE + Cr(VI) + CaCO₃. Over the experimental period, the enhancement of dissolved CaCO₃ was much greater than the iron passivation by secondary mineral precipitates. <br /><br /> Based on the laboratory experiments, Jeen (2005) developed an empirical formula relating the decrease in iron reactivity to the accumulation of secondary minerals, and incorporated this formula into kinetic expression of an existing multi-component reactive transport model (MIN3P). The same code was used in this study to simulate the experimental data. The model reproduced the observations from the columns in which TCE co-exists with Cr(VI) and CaCO₃ quite well, which suggests this model has applicability to predict the long-term performance of an iron PRB when treating groundwater containing Cr(VI), TCE and CaCO3, though there are some potential areas for improvements, including inconsistent volume fractions of secondary carbonate minerals and Fe(III)/Cr(III) products between experimental measurements and model simulation results, the reactive surface area concept, and inability to adapt changes in iron corrosion rates. <br /><br /> The long-term performance for a hypothetical scenario in using an iron PRB (40 cm thick) to treat groundwater where TCE (5 mg/L) co-exists with Cr(VI) (10 mg/L ) in the presence of CaCO3 (300 mg/L) was simulated. The simulation indicated that Cr(VI) was completely treated over a period of 30 years, however, TCE broke through before 20 years, and substantial porosity was lost due to the accumulation of carbonate precipitates. The prediction could be valuable in the design of PRBs or in the development of effective maintenance procedures for PRBs treating groundwater co-contaminated with chromate and TCE.

A Deep Geologic Repository (DGR) for Low and Intermediate Level (L&IL) Radioactive Waste has been proposed by Ontario Power Generation for the Bruce Nuclear site in Ontario Canada. The DGR is to be constructed at a depth of about 660 m below ground surface within the argillaceous Ordovician limestone of the Cobourg Formation. The objective of this thesis is to develop a regional-scale geologic conceptual model for the DGR site and to describe modelling using FRAC3DVS-OPG that provides a basis for the assembly and integration of site-specific geoscientific data. The numerical model is used to explain and illustrate the influence of conceptual model, parameter and scenario uncertainty on predicted long-term geosphere barrier performance. The modelling also provides a framework for hydrogeologic and geochemical investigations of the DGR, serves as a basis for exploring potential anthropogenic and natural perturbations to the DGR system, and demonstrates the long-term stability of the deep system. In the geologic framework of the Province of Ontario, the Bruce DGR is located west of the Algonquin Arch within the Bruce Mega-Block at the eastern edge of the Michigan Basin. Well logs have been used to define the structural contours at the regional and site scale of the up to 37 units that may be present above the Precambrian crystalline basement rock. The regional scale domain is restricted to a region extending from Lake Huron to Georgian Bay. While the selection of a larger domain might decrease the contribution of boundary condition uncertainty to any uncertainty in any site-scale performance measure, it significantly increases the contribution of the uncertainty in the spatial characterization to the uncertainty of the selected measure. From a hydrogeologic perspective, the domain can be subdivided into three horizons: a shallow zone characterized by the units of the Devonian; an intermediate zone comprised of the low permeability units of the Silurian and the shale units of the upper Ordovician; and a deep groundwater domain or zone characterized by units, such as the Cobourg formation, with stagnant water having high total dissolved solids concentrations that can exceed 200g/l. Hence, the conceptual model of the Bruce DGR site required the development of constitutive models that relate the fluid density and viscosity to the fluid total dissolved solids (TDS), temperature and pressure. The regional-scale hydrogeologic modelling will help demonstrate that at the proposed repository horizons, there are low energy gradients and that the combination of the low permeabilities and gradients will result in diffusional groundwater systems with favourable retardation properties.

Two municipal landfills and one public septic system in Southern Ontario were studied as potential sources of the pharmaceuticals ibuprofen, carbamazepine, gemfibrozil, caffeine, sulfamethoxazole, and naproxen to groundwater. The background chemistry at each site was also determined. Pharmaceutical analysis was conducted using isotope dilution techniques, coupled with solid phase extraction followed by high performance liquid chromatography electrospray tandem mass spectrometry (HPLC-ESI- MS/MS). An assessment of method performance and extensive quality assurance and quality control practices were employed. At the septic system site, pharmaceuticals were detected at the furthest sampling point, 30 m downgradient from the source area. The highest concentrations measured in groundwater were for carbamazepine (2,050 ng L-1), sulfamethoxazole (1,990 ng L-1) and ibuprofen (1,790 ng L-1). The other pharmaceuticals analysed were observed at concentrations in the range of <1 to 10 ng L-1 (gemfibrozil), <8 to 625 ng L-1 (naproxen), and <1 to 160 ng L-1 (caffeine). Under saturated groundwater transport, attenuation was not strong within the plume as all pharmaceuticals were detected at distance from the source. In the unsaturated zone, most pharmaceuticals appeared to be more greatly attenuated than in the saturated zone. This greater extent of removal in the unsaturated zone is attributed to increased degradation associated with elevated oxygen concentrations. At the two landfill sites, no pharmaceutical compounds were detected in any of the groundwater samples collected within previously defined plumes. Assuming these drugs are disposed in landfill wastes, the absence of detections suggests degradation and attenuation of these pharmaceuticals is occurring. Some of the conditions that may contribute to attenuation include a thick unsaturated zone, strongly reducing conditions, and high sorptive capacity of the waste. Specifically, waste typically has a higher organic content than aquifer materials, and a lower pH, particularly in the early stages of decomposition. These conditions would result in a potentially higher attenuation of drugs within the waste pile. This study suggests that management programs focused on protection of groundwater quality should take into consideration the potential persistence of pharmaceuticals in septic system environments.

Packer tests in boreholes in fractured rock involving injection or withdrawal of water in borehole segments have been standard practice in bedrock hydraulic investigations pertaining to geotechnical and water resource projects since the 1950’s. However in contaminant hydrogeology, the tests are conducted to assess groundwater velocity and contaminant fluxes and therefore, much improved resolution and measurement accuracy is needed. For this thesis study packer testing equipment was designed specifically for studies of contaminant behavior in fractured rock with the ability to conduct four types of hydraulic tests: constant head/flow injection step tests, slug tests, pumping tests and recovery tests, all in the same borehole test interval without removing the equipment from the hole while acquiring high precision data for calculation of transmissivity (T) and fracture hydraulic apertures (2b). This equipment records pressure above, within, and below the test interval to gain insights regarding open borehole flow patterns, and to identify short circuiting to the open borehole above or below the test interval. The equipment measures flow rates as low as 6 ml/min up to 20 L/min, and the temperature in the test interval and at the ground surface is measured to account for density and viscosity variations. Each type of test is conducted repeatedly over a wide range of imposed applied pressures and flow rates and the equipment was applied to assess performance of this new methodology for packer testing and gain new insights concerning fractured rock hydrology in 6 boreholes in the fractured dolostone aquifer underlying the City of Guelph, Ontario. In the first stage of the equipment application in the fractured dolostone aquifer, over 150 high precision straddle packer tests using constant rate injection (Q) were conducted to identify the conditions of change from Darcian (linear) to non-Darcian (non-linear) flow based on the Q vs dP relationship where dP is the applied pressure above ambient. In the Darcian regime, the linear Q vs dP relationship passes through the origin (0,0) where the ambient pressure represents static conditions (i.e. Q=0 and dP=0). After the onset of non-Darcian flow, proportionally less Q per unit dP occurs so that the interval transmissivity (T) calculated from the test results using Darcy’s Law based models is underestimated by as much as an order of magnitude. The Darcy-Missbach equation was found to be a robust conceptual model for representation of step constant Q tests in which the linear proportionality constant relates Qn vs dP. It was found that quantifying non-linear flow allows for a more accurate determination of the linear data to obtain better estimates of T and hence the hydraulic apertures derived from the T using the Cubic Law. In order to obtain hydraulic apertures from the packer test T values, the number of hydraulically active fractures in the test interval is needed. The only data collected regarding individual fractures was the core log created during the coring process and the acoustic televiewer log, both of which identify the location of fractures, but neither could tell if the fractures identified were hydraulically active. A sensitivity analysis concerning the effects of non-linear flow and the number of hydraulically active fractures on the calculated hydraulic aperture shows that the number of fractures selected as hydraulically active has the greatest effect on the aperture values. A new approach is proposed for determining apertures from hydraulic tests in fractured rock utilizing the onset of non-linear flow to aid in the choice of the number of active fractures present in the test interval. In the second stage of the equipment application, the four types of hydraulic tests (constant head, pumping, recovery, and rising/falling head slug tests) conducted in the same test interval at gradually increasing flow rates showed that non-linear flow can be most easily identified and quantified using constant head tests providing a higher degree of certainty that the data used to calculate T are from the Darcian flow regime. Slug tests are conducted most rapidly, but formation non-linear behavior is commonly exaggerated by non-linearity within the test equipment at large initial displacements. However, the equipment non-linearity can be accounted for using a Reynolds number (Re) analysis allowing identification of the non-linear flow in the formation. In addition, non-linear flow can interfere with evidence of fracture dilation. The pumping and recovery tests are the most time consuming because of the relatively long time required to reach steady state. However, these tests offer the most potential to give insight into the influences of the peripheral fracture network and rock matrix permeability on test results In addition to the actual transmissivity of the test interval T values obtained from packer tests can be influenced by several factors including non-linear flow in the formation and in the test equipment, aperture dilation or closure, hydraulic short circuiting or leakage from the test interval to the open borehole and dual permeability properties of the system (fractures and matrix). The equipment and procedures developed in this thesis provide an improved framework for identifying these influences and in some cases avoiding them so that the aperture values calculated from T measurements are more accurate than those obtained through conventional approaches. In the conventional procedures for packer testing in fractured rock as recommended in manuals and guidance documents, the applied head and flow rate can be expected, based on the results of this thesis, to produce transmissivity values biased low because of non-linear (non-Darcian) flow.

The importance of the hydroxyl radical (OH) to tropospheric chemistry is well known. The radicals’ ability to react with most atmospheric trace gases allows it to act as the main removal mechanism for these gases. Due to the highly reactive nature of OH, the oxidizing capacity of the atmosphere is often defined simply by its concentration. Owing to its significant role, knowledge of all OH chemistry, homogeneous and heterogeneous, is important to understanding the chemistry of the troposphere. In order to accurately predict future levels of OH and other trace gases, a thorough understanding of all processes and variables involved in the emission and sequestering of these compounds is essential. The gas-phase chemistry of OH is well known and has been extensively characterized. The one process that scientists are still trying to fully understand is its involvement in heterogeneous chemistry. Some studies have suggested that the inclusion of OH heterogeneous chemistry is important to fully model tropospheric chemistry, while other studies have suggested that it can be neglected entirely. It is therefore important to study and understand the conditions in which heterogeneous chemistry is significant. In order to do this accurately, scientists must first understand the process and magnitude of the uptake of OH onto atmospherically relevant surfaces. The main objective of this work was the development of a new analytical tool for the study of heterogeneous hydroxyl radical reactions. To this end, experiments were conducted to determine the most efficient approach to couple a low pressure aerosol flow tube (LP-AFT) to a chemical ionisation mass spectrometer (CIMS). The use of CIMS allowed for the accurate detection and quantification of hydroxyl radicals. Through iterative experimentation the system was designed and became operational. Experimental work focused on laboratory studies of reaction kinetics, with data reported in this work representing the reactions of OH with model atmospheric aerosols. The uptake of OH on organic aerosols was examined using the newly developed LP-AFT-CIMS system at standard temperature. Liquid oleic acid particles were used to mimic atmospherically relevant particles. The uptake coefficient, γ, on oleic acid particles was determined to be 0.49 ± 0.08 for a log-normally distributed aerosol at ~400nm. This value is in very good agreement with currently published data. However, the overall error of this method (~16%) is observed to be lower than other currently available methods, which have errors ranging from 20 – 30%. It is postulated that the mass accommodation coefficient, α, for OH radicals on organic surfaces approaches this value under standard atmospheric conditions. It’s also suggested that under the correct conditions the heterogeneous loss of OH could contribute to the overall budget of the OH radical. Further atmospheric implications of this reaction are discussed.

The current study monitored DO, stable isotopes of O2 (δ18O-O2 and δ18O-H2O) along with water quality parameters on both a diel and day-time only basis in the Grand River over several seasons and locations. A dynamic dual mass-balance model was developed to quantify rates of community respiration (CR), gross primary production (GPP), and gas exchange coefficients (k) in the Grand River, Ontario, Canada. Monitoring was conducted at three locations along a longitudinal gradient: 1) West Montrose, located upstream of the cities of Kitchener and Waterloo in a predominately agricultural landscape; 2) Bridgeport, located downstream from WM and the Conestogo tributary confluence, and 3) Blair, located downstream of the cities of Kitchener and Waterloo. Values of k in the Grand River ranged from 3.6 to 8.6 day-1, over discharges ranging from 5.6 to 22.4 m3 s-1. The k values were relatively constant over the range of discharge conditions studied. Overall, k values obtained with the isotope model for the Grand River were found to be lower than predicted by the traditional approaches evaluated, highlighting the importance of determining site-specific values of k. Metabolism results indicated that the Grand River is negatively impacted by both agricultural and urban inputs from the surrounding catchment. Metabolism rates in the Grand River (GPP = 2.2 to 19.9 and CR = 4.0 to 29.6 g O2 m-2 d-1) were found to be greater than published estimates for relatively undisturbed systems. In particular, net production at Blair was consistently below zero, indicating that DO inputs are not sufficient to overcome the oxidative demand upstream of this location. The addition of an isotopic mass balance provides for a corroboration of the input parameter estimates between the two balances, and constrains the range of potential input values to allow for a better estimate of GPP, CR and k. Input parameter uncertainty and sensitivity most likely reflect the dynamic processes occurring in the Grand River watershed. A better understanding of processes affecting δ18O-O2 would improve the capability of the model to replicate observed data, and provide more confidence in predicting metabolic processes in impacted rivers.

In seismology, the capability of an earthquake to induce other seismic events has been widely accepted for decades. For example, the term aftershock involves a strong relation of such a seismic event with the incidence of a main shock. Moreover, hydrological changes (water level in wells and streams, geyser eruption and remote seismicity) in response to remote earthquakes have been reported for many years. A matter of current debate concerns the spatiotemporal scale of interaction among seismic events. However, there appears to be no clear image of what is the exact method of transmission of the triggering energy for the phenomena listed above. It appears that the P-wave and the S-wave are inadequate in terms of ground strain magnitudes at teleseismic distances, while the amplitude of the surface waves generally decreases exponentially with depth in the Earth and could not be responsible for triggering deeper earthquakes or deep-seated fluid flow fluxes in 3-5 km deep reservoirs. This leaves some other wave as a possible triggering energy sources. <br /><br /> This thesis is based on a diffusion-dynamic theory that predicts a low velocity displacement wave, called a soliton wave, propagating in liquid-saturated porous media with velocity ~100-300 m/s, analogous to a tsunami that travels with the loss of little energy. This is hypothesized to be the mechanism for energy transfer that could be sufficient to promote changes in local pore pressure and therefore to alter the ambient effective stresses. It is also hypothesized that a soliton wave packet is emitted by a primary seismic event and may trigger sympathetic secondary earthquakes at a remote distance, fluid level fluctuation in wells, changes in geyser eruption behaviour, and changes in microseismic frequency, amplitude and patterns in appropriate places (e. g. under water reservoirs, in areas of active hydrothermalism, in tectonically active areas, and so on). <br /><br /> This thesis undertakes a review of some of these phenomena, and finds that the evidence as to what is the triggering mechanism is not clear. Also, it appears that the soliton hypothesis is not at all disproved by the data, and there may be some evidence of its existence. <br /><br /> To reveal the evidence of this kind of wave (soliton) in nature, real sequence and K-Q cases velocity data bases of earthquake interactions in the year of 2003 have been constructed by using information from Incorporated Seismological Research Institute (IRIS). The qualitative and quantitative analysis demonstrates that interactions between seismological and hydrological systems due to soliton waves are a definite possibility. However, the growth of fluid fluxes, geysers eruption and remote seismicity are controlled by both the principal stresses and the pore pressure. Hence, this interaction depends on the hydromechanical properties of rock such as permeability, compressibilities, and viscosities of fluids, saturations, and porosity. Perhaps the strongest argument in favour of a low-velocity soliton trigger is that the other seismic waves seem to be inadequate, and there is no evidence for their actions as a trigger. <br /><br /> The practice of detection and analysis of a soliton is not undertaken in this work. Because current devices are incapable to measure such a wave as they are on the surface and insensitive to liquid-solid coupling, sensitive and precise sensors in the low frequency range must be installed within the liquid saturated zone, preferably under the water table, to advance further work.

Iron permeable reactive barriers (PRBs) have been installed at sites contaminated with various reducible organic and inorganic chemicals, particularly chlorinated solvents, worldwide. Many geochemical factors can affect the performance of iron PRBs. Chemicals such as nitrate and Cr(VI) may act as competing oxidants when co-exist with chlorinated solvents. Previous studies observed declines in the rate of TCE degradation by granular iron in the presence of nitrate. Passive oxide formation on the iron surface and an increase in corrosion potential of the iron were determined to be the mechanisms for the decline. Cr(VI), being a stronger oxidant than nitrate, may have a similar but greater effect on the iron reactivity. In addition, chromium oxide as well as dissolved CaCO₃, a common groundwater constituent, form secondary precipitates and are likely to further affect the iron reactivity. The primary objective of this study was to determine the effects of Cr(VI) and dissolved CaCO₃ on the iron reactivity towards TCE and Cr(VI) reduction and to provide mechanistic explanations for the observation. In addition, the applicability of a modified reactive transport model (Jeen, 2005) to the system in which chromate and CaCO₃ co-exist with TCE was evaluated. <br /><br /> Column experiments, including measurements of corrosion potential and surface film composition using Raman spectroscopy were conducted. Five column tests were carried out with input solutions consisting of different combinations of TCE (5 mg/L), Cr(VI) (10 mg/L) and dissolved CaCO₃ (300 mg/L) for eight months. <br /><br /> The results from the column receiving only Cr(VI) showed that Cr(VI) was reduced rapidly by the granular iron and was not detected beyond 10 cm from the influent end of the column by the end of the experiment. However, Cr(VI) profiles migrated from the influent end further into the column overtime, suggesting progressive passivation of the iron near the influent end of the column. The gradual increase in corrosion potential (up to 180 mV positive shift) at the port 3 cm from the inlet with the migration of Cr(VI) profiles suggests the formation and accumulation of higher valent iron oxides, such as hematite and goethite, together with Cr(III) products on the iron surface, passivating iron material. Raman spectroscopic measurements confirmed the presence of passive iron oxides at the end of the experiment. For the column receiving Cr(VI) + TCE, the co-existence of TCE did not affect Cr(VI) reduction kinetics. However, the presence of Cr(VI) affected TCE degradation significantly. Two segments in the migration of TCE profiles are identified: the first segment near the influent end of the column, where the iron was still active towards Cr(VI) reduction but inactive towards TCE degradation, and the second segment where Cr(VI) was fully removed and the TCE degradation continued to follow pseudo-first-order kinetics. The migrations in Cr(VI) and TCE profiles suggest that iron was passivated by Fe(III)/Cr(III) products, and Cr(VI), being a stronger oxidant, was reduced much more rapidly than TCE. It is expected that the first segment of TCE profiles would extend gradually with the migration of Cr(VI) profiles over time. <br /><br /> When dissolved CaCO₃ was added to the columns with Cr(VI) and TCE, either as single contaminant, or as co-contaminants, the pH values near the influent end of the columns remained relatively low (~ pH 7), thus, the presence of dissolved CaCO₃ resulted in a stable corrosion potential and faster degradation rates of TCE and Cr(VI). Over time, however, Cr(VI) reduction and iron corrosion produced OH- and shifted the carbonate-bicarbonate equilibrium, resulting in the precipitation of secondary carbonate minerals, as detected by Raman analysis in the three columns containing CaCO₃. The precipitation and accumulation of the secondary minerals on the iron surface gradually decreased iron reactivity, as indicated from the progressive migrations of TCE profiles in the column receiving TCE +CaCO₃ and of second segment TCE of profiles in the column receiving TCE + Cr(VI) + CaCO₃. Over the experimental period, the enhancement of dissolved CaCO₃ was much greater than the iron passivation by secondary mineral precipitates. <br /><br /> Based on the laboratory experiments, Jeen (2005) developed an empirical formula relating the decrease in iron reactivity to the accumulation of secondary minerals, and incorporated this formula into kinetic expression of an existing multi-component reactive transport model (MIN3P). The same code was used in this study to simulate the experimental data. The model reproduced the observations from the columns in which TCE co-exists with Cr(VI) and CaCO₃ quite well, which suggests this model has applicability to predict the long-term performance of an iron PRB when treating groundwater containing Cr(VI), TCE and CaCO3, though there are some potential areas for improvements, including inconsistent volume fractions of secondary carbonate minerals and Fe(III)/Cr(III) products between experimental measurements and model simulation results, the reactive surface area concept, and inability to adapt changes in iron corrosion rates. <br /><br /> The long-term performance for a hypothetical scenario in using an iron PRB (40 cm thick) to treat groundwater where TCE (5 mg/L) co-exists with Cr(VI) (10 mg/L ) in the presence of CaCO3 (300 mg/L) was simulated. The simulation indicated that Cr(VI) was completely treated over a period of 30 years, however, TCE broke through before 20 years, and substantial porosity was lost due to the accumulation of carbonate precipitates. The prediction could be valuable in the design of PRBs or in the development of effective maintenance procedures for PRBs treating groundwater co-contaminated with chromate and TCE.

A Deep Geologic Repository (DGR) for Low and Intermediate Level (L&IL) Radioactive Waste has been proposed by Ontario Power Generation for the Bruce Nuclear site in Ontario Canada. The DGR is to be constructed at a depth of about 660 m below ground surface within the argillaceous Ordovician limestone of the Cobourg Formation. The objective of this thesis is to develop a regional-scale geologic conceptual model for the DGR site and to describe modelling using FRAC3DVS-OPG that provides a basis for the assembly and integration of site-specific geoscientific data. The numerical model is used to explain and illustrate the influence of conceptual model, parameter and scenario uncertainty on predicted long-term geosphere barrier performance. The modelling also provides a framework for hydrogeologic and geochemical investigations of the DGR, serves as a basis for exploring potential anthropogenic and natural perturbations to the DGR system, and demonstrates the long-term stability of the deep system. In the geologic framework of the Province of Ontario, the Bruce DGR is located west of the Algonquin Arch within the Bruce Mega-Block at the eastern edge of the Michigan Basin. Well logs have been used to define the structural contours at the regional and site scale of the up to 37 units that may be present above the Precambrian crystalline basement rock. The regional scale domain is restricted to a region extending from Lake Huron to Georgian Bay. While the selection of a larger domain might decrease the contribution of boundary condition uncertainty to any uncertainty in any site-scale performance measure, it significantly increases the contribution of the uncertainty in the spatial characterization to the uncertainty of the selected measure. From a hydrogeologic perspective, the domain can be subdivided into three horizons: a shallow zone characterized by the units of the Devonian; an intermediate zone comprised of the low permeability units of the Silurian and the shale units of the upper Ordovician; and a deep groundwater domain or zone characterized by units, such as the Cobourg formation, with stagnant water having high total dissolved solids concentrations that can exceed 200g/l. Hence, the conceptual model of the Bruce DGR site required the development of constitutive models that relate the fluid density and viscosity to the fluid total dissolved solids (TDS), temperature and pressure. The regional-scale hydrogeologic modelling will help demonstrate that at the proposed repository horizons, there are low energy gradients and that the combination of the low permeabilities and gradients will result in diffusional groundwater systems with favourable retardation properties.