| dc.description.abstract | The mining industry in South Africa is one of the primary sources of water pollution. Closed down mines leave a legacy of chemically polluted water after closure, and current operating mines are continuously discharging polluted water into the environment, consequently polluting water resources. The chemical pollutants found in this mine waste water include high concentrations of sulphate (up 5 000 mg/L), dissolved heavy metals and iron (II). This water can have pH levels as low as 2.5 and it is therefore named Acid Mine Drainage (AMD). AMD is formed as a result of the oxidation of pyritical material, which becomes exposed to oxygen and water during mining activities. While neutralization of the AMD and metal removal has been achieved by using alkaline compounds, sulphate removal to environmentally acceptable levels is still a source of controversy. Environmental, health and water governing bodies are exerting pressure on mines to treat water for sulphate content. A number of processes aimed at sulphate removal, to acceptable levels, are currently in use, and the scope of this thesis concentrates on their development and optimization. Previous research has shown that while limestone (CaC03) and lime (CaO) were conventionally used for neutralization and metal removal from AMD, these two chemicals also have a potential for partial sulphate removal from sulphate rich water, via gypsum (CaSCU) crystallization. However, a number of drawbacks such the inability to remove sulphate to low levels without addition of excess chemicals has led to the exploration of other chemicals for potential sulphate removal from AMD. Barium salts (Ba(OH)2, BaS and BaCOa) can also remove sulphate from sulphate rich water, stoichiometrically, via BaSCU precipitation. The use of these chemicals offers an added advantage of recyclability, via thermal reduction of the precipitated BaSCU to BaS, in the presence of a reducing reagent. In this thesis, the integrated barium carbonate process for sulphate and metal removal from AMD is presented and consists of the pre-treatment with lime, removal of sulphate as barium sulphate by dosing barium carbonate, the thermal reduction of BaSC>4 to BaS for sulphur production and possible BaC03 recycling, and finally H2S stripping from the concentrated solution of the recovered BaS is done, leading to sulphur production. From the beaker studies it became evident that sulphate can be removed by dosing the stoichiometrical amount of BaC03 into the sulphate rich water. The rate of sulphate removal is dependent on the BaC03 concentration and the sulphate removal is not directly inhibited by the presence of magnesium in the treated water, as was previously assumed to be the case. The sulphate removal rate is only retarded by an alkalinity > 200 mg/L (as CaC03.) An online particle size measuring experimental set-up was developed to study the precipitation process of BaSCU from the reaction of Ba+2(aq) and S04"2(aq), with the aim of enhancing particle growth over nudeation, in order to produce BaSC>4 crystals with improved settling properties. The studies have demonstrated that by changing the reactant concentration, number of feeding points into the precipitator and the stirrer speed, one can affect the extent of the feeding zone and the level of supersatu ration in this zone and control the size of the precipitated particles. The Crystal Size Distribution (CSD) analysis of the precipitation process has shown that growth takes place in all the crystal size ranges, but faster in larger crystals, suggesting a size dependent type of particle growth rate. By lowering the concentrations of the fed Ba+2 the local supersatu ration was reduced thus lowering the nudeation rate and as a result increasing particle growth rate. Increasing the number of feed points into the precipitator tube enhanced particle growth. Improved mixing due to increased stirrer speed led to increased particle growth rate, in all particle size ranges and reduced nudeation rate. However, an excessive high stirrer speed led to attrition type nudeation, the result of which is reduced/stunted growth of particles. The results from the studies, carried out using activated carbon as a reducing agent, in a furnace, have shown that the optimum temperature for the reduction of BaSCU to BaS is 950 - 1050°C, within 15 minutes for a complete reduction in a tube furnace. More than 1 hour was required for more than 60% yield to be obtained in a muffle furnace. The presence of CaCC>3 in the reaction mixture does not have a significant effect on the BaS% yield and the BaS% yield in the tube furnace is higher compared to the muffle furnace.
The TGA isothermal studies have revealed that the reduction rate of BaSCU using CO is dependent on the partial pressure of CO in the system and is also dependent on the temperature. A first order reaction rate with the average activation energy of 149 (±10) kJ/mol and constant value (k) of 0.59 were found to best describe the reaction. An effective H2S stripping depends on the balance between the C02 concentration and the sulphide concentration in the BaS solution. The molar proportionality between C02 fed and the sulphide stripped was almost equal to 1 only when the pH of the BaS solution was > 12. The results from these studies will be used in the pilot scale implementation of the integrated barium carbonate process which is currently underway at Harmony mine in Randfontein. | en_US |
