None

This thesis aims at geohydrologically assess arsenic in the Jagersfontein area and hence proposing remediation option. The field procedure carried out involved hydrocensus and sampling where samples were collected and analysed for micro and macro parameters with emphasis on arsenic. Isotopic (18O, D (deuterium) and Tritium 3H) analysis was also carried out to characterise the water in terms of its origin, fate and age. A total of 20 boreholes were sampled. Historical data from Bloemwater was also incorporated in the study. According to the chemical analysis, the most abundant cation is magnesium followed by calcium, and the most abundant anion is sulphate. Three different water types were classified; sodium/potassium sulphate, magnesium sulphate and calcium/magnesium bicarbonate. Hence they are from different sources. Sample collected from the mineshaft showed highest arsenic concentration which is higher than the South African National Standards (SANS 241, 2006) guideline for acceptable arsenic concentration in drinking water while all other individual boreholes in the main town of Jagersfontein showed low arsenic concentration (<0.006 mg/l); hence concluding that arsenic contamination comes from the mine. Mine water that has been treated in package plant (JFinal) still shows high arsenic concentration. Water from Kalkfontein dam is being treated in a bigger treatment plant (went operational in July 2012) and mixed with treated mine water. This reduces the arsenic concentration of water to an acceptable level before it is finally supplied to the community. Spearmanâs correlation coefficient calculated show a weak correlation between arsenic and other chemical parameters and so it is concluded that arsenic and these elements are not of same source implying the source of arsenic is not geogenic. Isotopic analysis shows the water samples plot very close to the global meteoric water line (GMWL) implying the groundwater is meteoric and is derived from the atmosphere. Clustering of the samples around the GMWL hints they might be of the same age. However one sample (surface water sample) plotted away from GMWL and it is attributed to effects of evaporation. The hardness of the water samples is generally high. To solve this problem, individual owners of boreholes add water softeners (chlorine, crystals) to soften the water. There are several techniques available for removal of arsenic from water in large conventional treatment plants. The basic principles of these technologies are based on physical-chemical processes such as oxidation, coprecipitation and adsorption on coagulated flocs, adsorption onto sorptive media, ion exchange, and membrane filtration. Most treatment options need the oxidation of As(III) to As(V) for efficient removal of arsenic. For the case of Jagersfontein, coagulation and filtration technology (FeCl3 as coagulant) will be the best option because it is very efficient in removing arsenic (V) which happens to be the dominant specie in the water. It is the cheapest technique compared to the others and it is also less sophisticated. However what is currently done in Jagersfontein to reduce the concentration of arsenic in water is dilution/mixing. Mine water treated in a package is mixed with treated water from Kalkfontein dam. Water (not arsenic contaminated) from Kalkfontein dam is treated in a bigger treatment plant which went operational in July 2012.

Australasian, European, Canadian and American geo-environmental specialists have long been involved with pit lake studies in the physical processes and dynamics as well as the geochemical aspects of these mitigation methods and events. The use of pit flooding as an environmental post-closure option has however been studied and used to a limited extent in Southern Africa and Africa to a whole, with some recorded cases mostly being by accident rather than a planned mitigation. This document is a written paper on a hydrogeochemical investigation on the feasibility of flooding an open pit platinum mine in the Bushveld Igneous Complex as a post-closure mitigation option. Various data sets and sources were compiled and processed and served as input data into a hydrogeochemical model of the expected impacts on the conceptual meromictic pit lake environment. Weathering and speciation models allowed the evaluation of the current system with thermodynamic and chemical reaction processes leading to the hydrochemical composition of the groundwater as we observe it in the field. Furthermore transient mixing simulations between groundwater, surface water and rain water was allowed to occur under various reducing and evaporative conditions. The modelled lake chemistry was then evaluated against recreational, agricultural and domestic water use standards to investigate the feasibility of the lake to be used post-closure. Final recommendations and mitigation methods were proposed with the flooding of the open pit as a post-closure option declared feasible. A final deliverable of the study was a system thinking and modelling template for hydrogeochemical modelling in various scenarios to guide other scientist through the process of modelling fluid-rock, fluid-waste and fluid-fluid interaction.

The remediation of sites contaminated by dense non-aqueous phase liquids (DNAPLs) continues to present a significant environmental challenge globally. Contributing to this challenge is the difficulty in locating source zones due to local heterogeneities in the sub-surface. Heterogeneities are significant in fracture rock environments, such as those found in South Africa, which together with the fluid properties determine the fate and transport of DNAPLs. This research is based on evaluating the effectiveness of combining traditional and novel source zone characterisation methodologies in order to delineate chlorinated hydrocarbon DNAPLs in a fractured rock environment. The research documents and evaluates the characterisation process followed in the application of various methodologies to an Investigation Site in South Africa. A site-specific conceptual site model is presented indicating the delineation of the multiple chlorinated hydrocarbon DNAPL source zones at the site. Additionally, a DNAPL source characterisation approach is proposed for application in fractured rock environments. This approach allows for the convergence of traditional approaches (such as drilling within a fixed grid) with more novel approaches (such as high resolution sampling and analysis). The pioneering use of ribbon NAPL samplers (FLUTeTM activated carbon technology membranes) in South Africa is documented in this research. In situ source zone characterisation using this technology in a fractured rock environment is shown to be successful in determining depth discrete fracture transmissivities and residual DNAPL zones that would have gone unobserved through methods such as direct observation and testing rock cores with hydrophobic dyes. The efficiency of this technology renders it ideal for future continued use in South Africa.

Winter (1999) succinctly, made a profound statement to the effect that understanding the basic principles of the interaction between surface water and groundwater is needed for effective management of water resources. Hence, the research investigation was aimed at identifying appropriate methodologies for assessment of surface water â groundwater interaction, thus enhancing the understanding thereof. The methodology used entailed a review of national and international literature on related previous and current models, systems and methods used in assessment and quantification of water exchange between groundwater and surface water. This was then followed by relevant case study analyses where distinct areas where chosen based on availability of relevant data and information by previous investigators. The findings were that various methods and classification systems are widely available but the applicability thereof under the South African conditions depends on the conceptual understanding of the area or system under investigation, availability of data and the basic assumptions associated with the particular model or method. The surface water â groundwater interaction cannot easily be quantified with confidence without requisite data available. The other finding is that use of multiple techniques to reduce uncertainties and to confirm or verify the existence or nonâexistence of the interaction is essential. Preferably, at least one method should be utilised to trace flow or qualitatively establish the water exchange while the alternative method is used for quantitative estimation of the interaction between surface water and groundwater. Some of the products emerging from this research investigation include decision tables for choosing applicable analytical method, applicable numerical method and the framework for guiding the selection of appropriate methodologies for assessing or quantifying the interaction between surface water and groundwater. Knowledge generated is applicable to water resource management, resource protection, water allocation and monitoring.

The suitability of different electrodes arrays used during electrical resistivity tomography surveys for geohydrological studies in the Karoo rocks is evaluated through theoretical considerations, numerical modelling and field surveys. The theoretical considerations predict that the Wenner array is sensitive to vertical changes in resistivity in the subsurface, while the Dipole-Dipole array is sensitive to horizontal changes in resistivity and the Schlumberger array is sensitive to both vertical and horizontal changes in resistivity. The theoretical considerations also show that the arrays with the strongest signal strength, the greatest depth of investigation and the broadest horizontal data coverage are the Wenner, the Dipole-Dipole and the Pole-Pole arrays, respectively. Twenty synthetic geological models, simulating dolerite dyke and sill intrusions, weathered zones, faults zones or bedding plane fractures and different types of geological boundaries, are used to evaluate the sensitivity, depth of investigation, signal strength, as well as horizontal data coverage of the Wenner, Schlumberger, Dipole-Dipole and Pole-Pole arrays. Responses of these numerical models to the L1-norm, L2-norm and to random noise are also investigated. Numerical modelling results indicate that the Wenner array is sensitive to vertical changes in the Earthâs subsurface resistivity, has the strongest signal strength, yet the shallowest depth of investigation and the narrowest horizontal data coverage. The Schlumberger array, sensitive to both vertical and horizontal resistivity changes in the subsurface, has high signal-to-noise ratio and better depth coverage compared to the Wenner array. The Dipole-Dipole array is seen to be sensitive to horizontal resistivity changes in the subsurface, has great depth penetration, but a low signal-to-noise ratio. Although the Pole-Pole array has the broadest horizontal data coverage and the greatest depth of investigation this array is the most vulnerable to noise contamination. The robust constraint inversion (L1-norm) should be selected for inverse modelling if the goal of the surveys concerns groundwater exploration since this algorithm produces models with sharp boundaries between different geological bodies. Conversely, the smoothness-constrained least-squares inversion (L2-norm) is appropriate when studying groundwater migration and contaminant transport because the diffusion boundary of a contaminant plume in the subsurface geology is likely to be associated with a smoother variation in the resistivities. The L2-norm inversion method gives optimal results under such conditions where gradual changes in the subsurface resistivities are expected. Two-dimensional electrical resistivity tomography surveys were carried out at three different sites near the city of Bloemfontein where dolerite dyke and sill intrusions occur. Results showed that the Wenner array is the most suitable electrode array to use for field surveys conducted across sill intrusions, while the Schlumberger array is the most appropriate for field surveys over dyke intrusions or sills intruded by dykes. The Dipole-Dipole array presented significant limitations for the field investigations due to its low signal-to-noise ratio. Results from the field studies conducted around Bloemfontein are in agreement with theoretical considerations as well as numerical modelling except the minor limitation of the Dipole-Dipole array (array recommended in theory by some authors). Results demonstrate that the two-dimensional electrical resistivity tomography surveys employing either the Wenner or Schlumberger arrays, in conjunction with the appropriate inversion technique, would be of great benefit to geohydrological studies in Karoo rocks, particularly for boreholes siting and during contaminant transport studies.

Although naturally occurring in groundwater, nitrate presence is generally attributed and regarded as a contaminant. Elevated nitrate levels pose a threat once converted to nitrite, causing methaemoglobinaemia. This investigation serves to establish the origin, quantify the extent, distribution, and elucidate possible amelioration methods of nitrate in ground and surface waters of a chrome mine in Kroondal, South Africa. Already existing surface water and borehole (groundwater) points were sampled and monitored from 2011 to 2012. An independant sample collection was conducted twice during the year of 2012. ICP and isotope (hydrogen, oxygen and nitrogen) analysis were employed to see nitrate level behaviour. In 2011 and 2012, nitrate level elevations varied from acceptable to unacceptable, with maximums of up to 81.51 mg/l in both ground and surface waters (well above the Department of Water Affairs and Forestry domestic requirements of 10 mg/l). Hydrofacies plots attribute water chemistry to: calcium/magnesium bicarbonate, opencast mine, low sodium high salt and unpolluted waters. Domination in Mg cations occurs as would be expected within Bushveld Igneous Complex aquifers. This is a result of the abundance within mafic and ultramafic igneous rocks. Increased Nitrate concentrations in surface waters are suspected to be primarily a result of runoff from waste rock dumps, high in nitrate explosive residues. In groundwater, increased concentrations are suspected to be primarily a result of high nitrate concentrations in explosive residues from underground blasting, highly mobilised once in contact with water. A secondary cause is seepage from the tailings dam, waste rock, as well as from surface water, high in nitrate concentrations. Possible remediation strategies identified include Reverse Osmosis, Eco-Dosing and Biological De-Nitrification.

One reason why groundwater, so often constitutes the main source of drinking water in many cities and towns around the world, is because it is frequently present in sufficient quantities at the point of demand. However, this seemingly advantage may sometimes be its greatest disadvantage, especially in situations where the groundwater occurs at shallow depths and the area overlying the aquifer is populated densely. This problem is particularly relevant in the present technological age with its vast quantities of waste that is often disposed in an uncontrolled manner. Such a situation occurs at Douala the economic capital of Cameroon in central Africa. The city not only host more than 80% of industries in the country, but also has the largest urban population of approximately 3 000 000 with a population density of approximately 350 persons per square kilometre, which continue to increase at a rate of approximately 120 000 migrants per year from the rural areas, while the groundwater level is very shallow and may sometimes rise above the soil surface, especially during floods, which occur not too infrequently. Although the pollution problem is not restricted to groundwater as such, it is aggravated here, because of the ancient belief that wastes are safely disposed of, if buried below the earthâs surface. It took disasters like Love Canal and the Price Landfill to discover the detrimental effects that this practice may have on the population living on or near polluted aquifers. Extreme care therefore should be exercised to prevent the pollution of any aquifer that may pose problems to living organisms or to try and restore a polluted aquifer threatening the natural environment. Groundwater pollution should therefore receive urgent attention when discovered. This thesis describes an attempt to develop a set of guidelines for the restoration of the groundwater resources at Douala, based on the relatively new technique of permeable reactive barriers for groundwater remediationâa technique that is also increasingly applied in the restoration of the Superfund sites in the United States of America. Modern attempts to clean up contaminated aquifers, relies heavily on the use of suitable computational numerical models. Such models have in the past always been based on the classical hydrodynamic dispersion equation. However, an analysis of the equation in this thesis has shown that the equation cannot account for the longâtail contamination plumes characteristic of fractured rock aquifers. Fortunately, it is not too difficult to develop a more suitable equation. For, as shown in the thesis, all that one has to do is to replace the ordinary derivatives in the classical equation with fractional derivatives. Mechanistic modeling of physical systems is often complicated by the presence of uncertainties, which was in the past usually neglected in the models used in the restoration of aquifers.While these uncertainties have regularly been neglected in the past, it is nowadays imperative that any groundwater model be accompanied by estimates of uncertainties associated with the model. Although a large number of approaches are available for this purpose, they often require exorbitant computing resources. The present investigation was consequently limited to the application of the Latin Hypercube Sampling method applied to an analytical solution of the hydrodynamic dispersion equation. It has been known for years that the hydrodynamic dispersion equation discussed in Chapter 5, is not able to account for the longâtail plumes often observed in studies of contaminated fracturedârock aquifers. An approach frequently used to account for this is to replace the ordinary spatial and temporal derivatives in the hydrodynamic equation with fractional derivativesâa procedure confirmed in this thesis.

Enrichment of nitrate in groundwater is a worldwide phenomenon, mostly resulting from anthropogenic activities in densely populated areas. The objective of this study was to analyse the groundwater quality in Thaba Nchu in the vicinity of contamination sites, where the nitrate contamination of groundwater has been increased along the last decades. This is important to control and manage groundwater quality of aquifer systems in the rural areas. A multi-approach methodology was followed. During this approach, the hydrogeochemistry of major constituents (both hydrodynamic and hydrochemical parameters), as well as the concentrations of environmental isotopes (18O, 2H) and nitrate isotopes (15N and 18O) were used to provide information on land use in order to trace sources of waters and solutes, and to study associated processes in shallow groundwater systems of the Thaba Nchu rural villages. According to the spatial distribution of nitrate contents, nitrate concentrations exhibit very high concentrations in BAL2, GL1, GL2, GL3, NP2, NP3 and SP. 65 % of the sampled wells exceeded the value of 6 mg/L as NO3-N. On-site sanitation in the study sites were the main cause of serious nitrate contamination given the superimposition of high nitrate concentrations with the distribution of on-ground nitrogen loadings. A connection of nitrate concentrations to rainfall conditions was found: High nitrate concentrations were recharged under drier conditions while lower nitrate concentrations appeared to be recharged under much wetter conditions. From the Piper diagram calcium chloride/nitrate water type showed significantly higher NO3-N concentrations (NO3-N > 100 mg/L) than the other water types. The Expanded Durov diagram showed a range of water types from fresh, clean water to mixtures from different sources. Nitrate has a less coherent distribution associated with high δ18O values, clearly suggesting either a non-conservative behaviour or more than one source. Different δ18O-NO3 trends suggested isotopically distinct, non-point source origins which varied spatially and temporally, due to different degrees of evaporation/recharge and number of on-ground nitrogen loadings. The plot of δ15N versus δ18O values indicated that animal and human wastes were the predominant NO3â sources, and a few boreholes from ammonium fertilisers and organic soil matter. KOM was the only borehole experiencing denitrification. A management strategy was developed consisting of a situation assessment, immediate actions, medium and longer-term actions. The management and reduction of groundwater nitrate levels depend on an understanding of the nitrogen sources and the pollution and nitrification mechanisms.

Not available