Research into the use of 'algal' biomass for human consumption is receiving increased attention due to their favourable nutritional value, photosynthetic efficiency, and lower requirement of land and fresh water as compared to terrestrial crops. The Spirulina species, also known as Arthrospira, is of particular interest due to its high protein content and nutritional value. Open raceway pond systems are popularly used for commercial industrial scale cultivation of microalgae due to their economic feasibility. These open cultivation systems are, however, susceptible to contamination by other microorganisms. This raises concerns relating to suitability for human ingestion and the need to control bacterial growth to prevent contamination by pathogens and to minimise the overall bacterial load. Further, bacterial contamination in processed (harvested and dried) Spirulina biomass has been reported, suggesting that some of these contaminants may end up in the market ready product where appropriate processing approaches are not used. This study sought to identify the microorganisms that typically contaminate Spirulina cultivation ponds, to understand their interaction with Spirulina biomass during cultivation and to evaluate the vulnerabilities of these contaminants, in order to generate strategies for controlling their populations during open pond cultivation. The main objectives of this study were therefore: • To quantify the bacterial load in processed Spirulina powder from a single pilot facility to ascertain the presence of the contaminant in the final product derived from the outdoor pond system used as a case study, and to quantify the bacterial load in the outdoor cultivation cultures. • To identify and characterize the bacteria associated with these Spirulina cultures and processed powder from a pilot operation carried out in Franschhoek, South Africa, with a particular focus on evaluating the likelihood for pathogens. • To establish the dynamics of the relationship between Spirulina and bacterial growth under different environmental conditions including pH, salinity and temperature. • To develop practical methods to control and minimize contamination.

Maximising performance of microbial processes, including yeast-based processes, in an industrial setting requires understanding of the impact of process stresses. These may be the result of process configuration, dilution, temperature changes, hydrodynamic conditions or process perturbations. Methods to determine the microbial metabolic response to such stresses have long been sought, but are typically limited, often requiring the use of a suite of methods to assess the physiological status and state. The recent technical advances in microcalorimetry suggest potential for the use of isothermal microcalorimetry (IMC) to determine yeast viability and vitality and is investigated here. IMC is a laboratory method whereby the real-time heat produced by a chemical, biological or physical process is measured in the micro to nano watt range. It is proposed that this heat production may be correlated to the physiological state of the microbial catalyst and can be used to measure the impact of different stresses. In this study, the potential of IMC as a method for exploring process stress is investigated using Saccharomyces cerevisiae and its application in the beer brewing industry as a case study. Here, it is well known that yeast viability and vitality have commercial significance. IMC is sufficiently sensitive to detect the heat given off by 1000 yeast cells. However, IMC cannot distinguish between different heat flows within a system i.e. it is non-specific. The literature demonstrates how IMC has been used in the study of numerous microbiological fields, including the growth and metabolism of yeast. Previous studies have successfully derived the specific growth rate and cell numbers of a growing yeast population from analysing power and heat curves. The specific growth activity and specific growth retardation of yeast and how these parameters relate to bactericidal and bacteriostatic effects has also been examined by a number of authors. The key objectives of this study were to determine the viability and vitality of Saccharomyces cerevisiae using IMC and to assess the impact of stresses on yeast viability and vitality. This was achieved by measuring the thermal power produced by a growing yeast suspension as a function of its overall growth and metabolism. Two industrially relevant stresses were examined: cold shock and ethanol shock. The effect of these stresses has yet to be studied using microcalorimetry. The growth of Saccharomyces cerevisiae under ethanol stress was used as an inhibition study to isolate its effects on the growth thermogram. Following the generation of thermograms under control and stress conditions using IMC, a method for their quantitative analysis was developed. Curves were fitted to the heat data using an exponential growth equation and the time for the heat flow curve to peak was determined. From the exponential curve, the specific growth rate of the yeast was determined with a high degree of repeatability. The coefficient of the exponential term in the growth equation gave highly reproducible and distinguishable results relating to the viability and vitality of the initial yeast population. The time of peak heat flow was also affected by the initial viability and vitality of the yeast and was used to estimate the initial active cell population size.

This Thesis is embargoed. Only the Abstract is available

Cyanide heap leaching had been proposed as an alternative to the classic crush-mill-loatsmelt-refine route for processing platinum group metals (PGMs) from the Platreef ore body. Overall the process includes two stages of leaching. The first stage involves the thermophile bioleaching of the base metal (BM) sulphide minerals and acts as a form of pre-treatment to oxidise sulphur compounds and recovery valuable metals such as Cu, Ni and Co. The second stage focuses on cyanide-based heap leaching for the recovery of precious metals (PGMs +gold) from the solid residue of the first stage. Exploration and optimisation of this second stage in the context of a whole ore Platreef material is the focus of the present study. The first part of the study used a series of laboratory tests simulating heap leaching, conducted on coarse ore. The initial tests showed high recoveries of base metals (Cu, Ni and Co) could be achieved in a pre-treatment bioleach process, while in the second stage cyanide leach high levels of Pd and Au were extracted, but only 58% of the Pt after 60 days from the whole ore. It was observed that during the 60 day leaching period the rate of Pt leaching decreased considerably after 35 days. From the trajectory of the Pt leach curve from the 35 day mark onwards, it was observed that the leaching would not cease even after 60 days but would likely proceed but at that slow pace which indicated further Pt extraction would not be commercially viable in the long run. Mineralogical analysis has indicated that a significant component of the Pt in the ore is in the form a mineral sperrylite (PtAs2), which appears to leach slowly in cyanide as compared to other mineral forms such as certain tellurides and sulphides in the ore. Subsequently, efforts were made to investigate methods to improve the second stage leach process, in terms of Pt leaching from sperrylite, through further work on a pure mineral sample. The key focus was on finding a suitable oxidant that can be used in cyanide solutions, from among air, oxygen and ferricyanide, to facilitate the dissolution. Various tests using sperrylite mineral samples micronized to 5 μm in batch stirred tank reactors (BSTR) at 50°C were conducted. It was found that a combination of ferricyanide with cyanide extracted as much as 16 times more Pt than tests using only cyanide. The presence of air or pure oxygen did not contribute significantly to the amount of Pt leached in this system and made no difference at all in the leach tests using only cyanide. Further bench-scale studies focused on characterising the leaching mechanism of sperrylite in cyanide-ferricyanide solutions. It was found that the reaction, after proceeding at appreciable rates initially, tended to cease after 1 day, indicating some form of surface passivation, tentatively related to some form of solution equilibrium being achieved. However after re-leaching the sample with fresh solution, the Pt dissolution improved tremendously. This was further investigated in continuous leaching of a sample of the mineral using a small bed of sperrylite fixed in mini-columns. The results from the minicolumns showed the same leaching pattern as the experiments using BSTRs. It was eventually revealed that a suitable wash of the sperrylite sample using water removes the inhibiting layer and facilitates further and improved leaching. Unlike the cyanide-only system where the passivation was attributed to As build-up at the surface, in the cyanide-ferricyanide system it was attributed to adsorption of unknown reaction products on the mineral surface. Residual samples from batch leach experiments were analysed using X-ray photoelectron spectroscopy and showed samples from the cyanide-ferricyanide tests had less As on the surface than the untreated sample and the sample leached in cyanide. To some degree this supported the hypothesis that Pt leaching is eventually hindered by As passivation in a cyanide system. The presence of ferricyanide serves to oxidise As and thereby release more Pt in solution. Additionally, electrochemical techniques using a sperrylite electrode were employed to further understand the redox reaction under varying oxidation conditions. While the tests indicated a weak current under mildly oxidising conditions in cyanide solutions, this became rapidly limiting at potentials expected in a ferricyanide solution, indicating a form of surface passivation. An attempt was made to determine the number of electrons transferred during Pt dissolution to indicate the primary reaction mechanism through a long-term test held at constant potential, but dissolution rates were too small to be conclusive. Hence the study has shown that the cyanide-based heap leaching of PGMs from Platreef type ores is feasible in principle, but the dissolution of PtAs2 remains limited. While the study has given valuable pointers to understanding this observation, the conclusion is that PtAs2 is refractory in the given context and further development of this process remains promising through further investigation into the use of the cyanide-ferricyanide combination.

Mineral beneficiation processes such as base metal and coal mining produce large amounts of waste rock and coal discards that contain significant quantities of sulphide minerals with Acid Rock Drainage (ARD) generating potential. ARD is caused by the exposure of sulphide minerals, primarily pyrite (FeS2), to both water and oxygen, and microorganisms. This is a naturally occurring process, but the exposure of the sulphide containing mining wastes greatly accelerates ARD formation. Thus, ARD is a major issue associated with inactive mines, waste rock dumps and tailings impoundments, which over time presents a major environmental risk. The desulphurisation of coal discards, mine tailings and finely divided waste rock prior to their disposal has been proposed as a method of preventing ARD formation. This involves the selective separation of residual values from the waste rock, followed by selective separation of sulphide minerals – especially pyrite – from the residual waste material using a two-stage froth flotation to obtain a values stream, a low volume sulphide-rich concentrate that can be easily contained, and a high volume benign tailings fraction that can be safely disposed of. The technical feasibility of this two-stage process has been demonstrated; however, the cost of the flotation reagents used in this process are particularly high in comparison to the other operating costs, contributing as much as 75% of the operating costs for desulphurisation of coal fines. Furthermore, apart from being expensive, many of the inorganic flotation reagents are relatively toxic and could be hazardous to the environment due to their slow degradation rate. Microorganisms and their metabolic products have been identified in literature as potential reagents that can be used in the selective separation of sulphide minerals using froth flotation. Just like conventional chemical flotation reagents, the microorganisms assist separation through surface chemical alterations that modify a mineral’s hydrophobic properties, thus facilitating bioflotation. The aim of this study was to investigate the prevention of ARD formation through the desulphurisation of pyrite-containing coal discards and base metal hard rock samples using microbial cultures as alternative bioflotation reagents. In this study the feasibility of using P. polymyxa, R. palustris, R. opacus, B. subtilis, and B. licheniformis as biocollectors for the removal of pyritic sulphur in the second stage of the two-stage desulphurisation froth flotation process was investigated. Microbial screening tests were performed using a pyrite concentrate to assess each microbial culture’s affinity to pyrite and their ability to float the mineral in a batch flotation cell. Attachment experiments and batch bioflotation tests were carried out to screen for a microbial culture that showed potential. Following attachment experiments at pH 4 and pH 7, all microorganisms except B. licheniformis exhibited attachment to pyrite. The level of attachment was different for each microbial culture. P. polymyxa had the highest percentage attachment of 95.6 ± 1.0 % at pH 4 and 97.1 ± 0.7 % at pH 7 after 20 minutes of interaction. Subsequent results from the pyrite-only bioflotation tests revealed that R. opacus, R. palustris and B. subtilis did not affect the floatability of pyrite. P. polymyxa, however, showed a significant effect on the floatability of pyrite, achieving a cumulative mass recovery of 7.0 ± 0.42 % at pH 4 and 81.3 ± 0.4 % at pH 7. Zeta-potential tests revealed that P. polymyxa had the most neutral net surface charge across the pH range tested, while the other microorganisms had a large net positive or negative charge. Based on this result, it was deduced that the hydrophobicity of P. polymyxa as a consequence of its near neutral surface strongly made it seek out a surface to attach to rather than remaining suspended in water. Hence, P. polymyxa was chosen as the bio-collector candidate for the bioflotation separation of pyritic sulphur from coal discard and base metal hard rock samples. Despite the positive batch pyrite bioflotation tests, P. polymyxa was not successful for the flotation of pyrite from the coal discards nor did it upgrade pyritic sulphur to the concentrate, with the bioflotation results not significantly different from the negative control without collector. P. polymyxa did affect the floatability of the base metal hard rock, achieving cumulative mass recoveries comparable with the chemical control using PAX. However, there was no significant upgrade of pyritic sulphur content, with the biofloat achieving 22.6 % total sulphur in the concentrate which was significantly less than the 66.4 % total sulphur recovered with PAX. The study thus yielded positive results from fundamental studies of P. polymyxa’s ability to enhance the flotability of pyrite. However, tests using actual samples were less successful. Although P. polymyxa enhanced the floatability of the base metal hard rock, it did not achieve the aim of obtaining a low volume sulphide-rich concentrate as the PAX did. Recommendations for the continuation of this work include contact angle measurements and FT-IR spectroscopy to better understand the effects of P. polymyxa attachment, as well as performing a kinetic study on the growth of P. polymyxa alongside adaptation of the microbial culture to a pyrite mineral concentrate in order to test if this can improve selective flotation of the desired mineral owing to modified surface properties.

The world sugar price is constantly changing in response to supply and demand and is currently very low as compared to the prices it is sold at domestically in South Africa. The drop in the worldwide price of sugar is due to its oversupply as yields of sugar production have increased in recent years and subsidies and protection measures in other producing countries. The low prices also mean imports are cheaper than local sugar. This pushes down the average sugar price and leads to a low profit margin. Further, sugar production in South Africa is facing a number of challenges. The industrialization of the sugar belt in KwaZulu-Natal has resulted in less plantations and challenging topography for these. Incentivisation of small, medium and micro-scale commercial operations has increased the number of smaller scaled operations, with less economy of scale and less capital backing. Climatic factors have impacted crop yields. Production costs have increased in accordance with South Africa’s consumer price index whereas selling price has moved with the less inflationary global platform. Together, these have made the industry less economically viable. This has led to a need for value addition to sucrose and to eliminate the dependency on a single commodity. Re-positioning of sugar into value-added products has potential to boost the country’s economy by introducing other sources of revenue. Moreover there is a worldwide need to find alternative means to produce petroleum-based fuels and chemicals and bio-based products are being targeted to meet some of this need. A review of the global status shows that there has been value addition in the sugar industry producing mostly ethanol and other commodity chemicals such as surfactants, organic acids and polyols. It is therefore imperative to find sustainable ways of generating value added platform chemicals from sucrose. The quantitative and qualitative study of this project looks at determining the chemicals that should be considered as having the highest potential for value addition from sucrose in a South African context. The project was scoped to focus on chemicals and fuels that can be produced by biological conversions of sucrose. For the quantitative study, a set of 39 chemicals was selected from major studies performed globally on potential bio-based platform chemicals and these catalogued according to a set of criteria. The decision of the chemical/fuel to be studied was based on the gap in the chemical industry. This list comprised of chemicals that were selected in the US department of energy top 10 list in 2004 and 2010 and top 15 chemicals in the EU list in 2015. In addition to these, chemicals that are currently of interest (which were mostly chemicals that can be used as polymers and biofuels) were included to make up the list of 39 chemicals. The selected chemicals then went through a knock out selection where chemicals that cannot be produced with current technology from sugar or via a biological route were eliminated from the list. A quantitative analysis was then done on the remaining chemicals from the knock out stage. A weighting method which considered a series of factors was used to determine the top platform chemicals. The factors used were to identify platform chemicals that are at a high demand (both in South Africa and internationally), chemicals that showed great potential for profitability based on cost, technology readiness level and product yield. The quantitative analysis allowed seven chemicals to be selected. Finally a qualitative study based on interviews with experts in the field was done. Most of this information provided by the experts was supported by several literatures (Taylor, et al., 2015; Villadsen, et al., 2011; Choi, et al., 2015; Jansen & van Gulik, 2014). The qualitative study identified Succinic acid, Lactic acid and Citric acid as the top three chemicals. A techno-economic study was done on succinic acid, one of the most promising platform chemicals identified. The reasons for its selection was because it has a higher performance and it generates less carbon footprint than petroleum based succinic acid, competiveness for niche market, multiple application via BDO and PBS and its overall favourable environmental process that uses up carbon dioxide from the environment. Firstly, the succinic acid process was designed to be produced using Saccharomyces cerevisiae in a dual phase fed batch fermentation process. The overall design for the succinic acid process was based on the design proposed by Efe, et al (2013). A cost evaluation was then done on the design for an economic analysis. The economic analysis was done on the process to ascertain that there is indeed value addition of sucrose to the platform chemicals chosen. This was done in the form of profitability analysis of the process. An economic analysis of the design shows that the plant is profitable after the first year of operation. The total investment on the plant is R 22.3 billion and the start-up expense is R 1.05 billion. This project serves as a preliminary paper based overview of the general background for the selected platform chemicals that will be researched further in subsequent research.

Heap bioleaching is a hydrometallurgical technology, used to facilitate the extraction of valuable metals such as copper, gold, nickel and uranium from low-grade, typically sulphidic, ores. The process is highly complex as it is influenced by interactions of different sub-processes including flow of leaching solution around the ore particles, mass and heat transfer within and around the particles, chemical reactions, microbially-mediated reactions and microbial growth. Contact of leaching solution with mineral grains is necessary for oxidation of the sulphide minerals. However, a large fraction of the mineral grains is positioned below the surface of the ore particles and so contact with the liquid occurs through cracks and pores in the ore connected to the surface. Long extraction times and low metal recoveries typical of heap systems can be attributed to the slow leaching rate of these non-surface mineral grains as well as constraints on their accessibility. Most of the valuable grains that remain in the residue ores are non-surface grains. Therefore, investigation of the mechanism and behaviour of non-surface grain leaching and quantification of the factors contributing to their leaching is expected to be highly beneficial in the optimisation of leach conditions and recoveries. Non-surface grain leaching within large particles cannot be investigated via traditional experimental methods reliant on bulk measurements, 2D or destructive methodologies. However, it can be studied using high resolution, non-destructive 3D X-ray micro-Computed Tomography (μCT), an imaging technique for investigation of internal structure of opaque objects. X-ray μCT has previously been developed and used for investigation of different aspects of heap leaching. In the current study, the viability of using X-ray μCT to study heap bioleaching systems and affecting variables is assessed. This required establishment of procedures for measurement and analysis of sulphide and oxide mineral recoveries and leaching penetration distances. The feasibility of studying biotic heap leaching by X-ray μCT was explored through investigation of the relative energies required for high mineral resolution and avoidance of microbial inactivation. Specific bioleaching operating variables that were subsequently considered included: the accuracy and representivity of the X-ray μCT images, the influence of agglomeration pre-treatment, operating temperature, and type of ore on non-surface grain leaching. Addition of surfactants to the leaching solution was explored with the aim of changing surface activity to influence the penetration of the leach agent into pores and cracks in the ore. The effects of operating conditions on non-surface mineral grain leaching was studied using mini-column experiments. Three different low-grade ores, namely a chalcopyrite-rich ore, a malachite ore and a waste rock containing pyrite were prepared for the leaching experiment. The ores were crushed using a jaw crusher and comminuted down to 100% passing 16 mm. The products were sieved into six fractions (<0.25 mm, 0.25 - 1 mm, 1 - 2 mm, 2 - 5.6 mm, 5.6 - 8 mm, 8 - 16 mm) and each fraction then representatively split into smaller portions using a rotary splitter. One portion of each size fraction was taken for XRD, AAS and QEMSCAN analyses. Mini leaching columns were designed and constructed based on the target mineral grain distribution in the ores to ensure that the mineral grains were detectable using X-ray µCT, given its resolution limitations. The columns were charged with 50 g of agglomerated or non-agglomerated ore and lixiviant was provided at a flow rate of 2.55 mL h -1 for a period of 5.5 months for chalcopyrite and pyrite and 26 days for malachite in incubators at 30 °C, 37 °C and 65 °C. In order to select a surfactant suitable for use in a biological leach experiment, the effect of five different types and concentration of non-ionic surfactants on bioleaching microorganisms was studied in terms of microbial growth, ability for ferrous ion oxidation and chalcopyrite bioleaching. This was done in shake flask experiments using mineral concentrate. Based on the results of these experiments, Tween® 20 (10 mg L -1 ) was selected to study the effect of surfactant on non-surface mineral grain leaching in the mini-columns. Each column was scanned by X-ray μCT at 100 kV and 150 mA using a 0.38 mm copper filter and at a distance of 59.40 mm between X-ray gun and specimen. The advanced 3D analysis software Avizo® 9 was used to visualize and analyse image data. The Interactive Thresholding function in Avizo® 9 software was used for segmentation of ore particles from air and sulphide minerals from air and gangue minerals, to measure the target minerals' volume reduction during leaching. The Distance Map Algorithm was applied on a binary (segmented) image to calculate the distance of the sulphide mineral from the ore particle surface. Imaging of the whole mini-column was done before leaching and at the end of each experiment and imaging of certain sections was done at select time points during leaching to track temporal leaching dynamics. Good agreement was seen between the bulk mineral recovery data, determined using standard chemical assays, and the leaching curves generated using the X-ray µCT images for all the ores, confirming that the X-ray µCT images were a good quantitative measurement of the sulphide and oxide mineral leaching. Liquid microbial culture experiments were used to confirm that exposure to X-ray does not affect microbial activity for energy doses between 35 and 90 kV at 200-280 μA. However, X-ray exposure was found to have a slight negative influence at higher voltages of 120 and 150 kV, temporarily reducing the specific ferrous ion oxidation and suppressing the specific growth rate of the bioleaching microorganisms. The X-ray exposure thus negatively affected both the total microbial population available for leaching (population viability) as well as the metabolic activity of the individual microorganisms (population vitality). The effect of X-ray exposure on bioleaching cultures attached to a mineral surface was examined using pyrite-coated glass beads packed into mini-columns. The energy dosage limits identified in the liquid culture experiments were found to be compatible with the X-ray μCT imaging conditions (minimum energy dosage and sample position) required for acquisition of complete and accurate images of the columns at a resolution that allows identification of individual mineral grains. Following X-ray exposure, the performance of the exposed bioleaching mini-columns was equivalent to the unexposed control column. Similarly, the microbial activity and presence on the mineral surface appeared unchanged. Finally, the experiment was performed on the chalcopyrite ore and the microorganisms were found to still be able to convert Fe2+ to Fe3+ after 2 scanning runs. Thus, all sets of results confirm that X-ray μCT can be compatible with heap bioleaching experiments, while still permitting appropriate resolution of the mineral grains to make an X-ray μCT investigation worthwhile. However, cognisance that an upper limit of tolerable X-ray exposure exists must be taken. This may present a challenge if it is desired to image larger or denser ore samples which require a greater X-ray energy level for sufficient penetration of the sample by the X-rays and hence accurate imaging. In chalcopyrite leaching, increasing temperature from 37 °C to 65 °C resulted in clear enhancement of leaching based on both analysis methods, with the copper recovery increasing from 20% to 64% by the end of the leaching period, and the overall sulphide mineral dissolution increasing from 24% to 67%. Increasing temperature from 37 °C to 65 °C resulted in an increased leaching penetration distance and crack development in the particles, and thus an enhancement in copper recovery and sulphide mineral dissolution. This was in addition to the thermodynamically expected increased leaching rate. The maximum leaching penetration distance, beyond which no mineral volume change is observed, at 37 °C was 1.7 mm. This increased to 2.5 mm at 65 °C. As a result of addition of 10 mg L-1 Tween® 20 into the leaching solution, the final copper recovery was improved by 4% to 68% and the maximum penetration distance increased to 2.9 mm. However, when the availability of sulphide mineral was not rate limiting, the copper recovery and sulphide mineral volume reduction in the mini-column with surfactant was lower than the system without surfactant. This may have been due to depression of diffusion of ferric ion to the ore surface as a result of the formation of an adsorbed surfactant layer on the mineral surface. The performance with surfactant became superior as the amount of readily leachable mineral became limiting. In the pyrite waste rock, an increase in temperature did not have any effect on the maximum penetration distance and any increase in iron recovery was only for thermodynamic reasons. Similarly to the chalcopyrite ore, during the later period of leaching when readily exposed mineral grains have been depleted, the system performed better in the presence of surfactant. The addition of surfactant increased the maximum penetration distance from 2.7 to 2.9 mm. The cumulative copper recovery of 86% was obtained for malachite ore in 26 days of acid leaching and the maximum penetration distance was 2.2 mm. This study thus demonstrates the value of the X-ray µCT technique for quantitative investigation of non-surface mineral grain leaching and confirms that the maximum penetration distance can be affected with changing operation conditions or ore type. This study thus demonstrates the X-ray µCT technique for quantitative investigation of non-surface mineral grain bioleaching and confirms that the maximum penetration distance can be affected with changing operation conditions. Critically, the results confirm that X-ray μCT can be compatible with bioleaching microorganisms, while still permitting appropriate resolution of the mineral grains to make an X-ray μCT investigation worthwhile.

Acid rock drainage (ARD) is a growing concern, particularly in South Africa, as a country already classified as water scarce. ARD is defined as water that has been impacted by mining activities and typically has high levels of sulphate and heavy metals, at acidic pH. Similarly, high sulphate neutral rock drainage is of increasing concern. High sulphate content increases water salinity leading to adverse effects on human health as well as agriculture. Types of ARD and neutral rock drainage can be categorised into those that are produced in high volumes from groundwater rebound, and those that are generated from diffuse sources, as low-flow ARD. Low-flow ARD and neutral rock drainage are amenable to biological treatment of the sulphate component using sulphate reducing bacteria (SRB). Biological sulphate reduction (BSR) generates sulphide, which requires further treatment to remove it from the stream in gaseous form or as a solid sulphur-containing compound. Alternatively, it can also be used, in part, to precipitate metals present within ARD waters. Key challenges associated with SRBbased bioremediation include the cost of the supplemented electron donor needed for SRB to reduce sulphate, as well as the downstream management or treatment of the excess sulphide remaining. This investigation aimed to optimise a semi-passive treatment process which integrates BSR, and concomitant partial oxidation, by sulphur oxidising bacteria (SOB), of the sulphide produced to elemental sulphur. This is generated as a floating sulphur biofilm (FSB). These processes occur simultaneously within a linear flow channel reactor (LFCR), facilitating both treatment of the water stream to a fit-for-purpose water product, and recovery of the sulphur for use within fertilisers or fungicides. The work focused on the effective utilisation of electron donors and a sustainable option thereof, as well as the optimisation of partial sulphide oxidation and sulphur recovery. In addressing the cost of a supplemented electron donor, the use of a waste product is of interest. To explore this and build on earlier work within the Centre for Bioprocess Engineering Research (CeBER) labs, this study investigated the efficient use of the volatile fatty acids (VFAs) acetate, propionate and lactate. Firstly, the study investigated propionate, a common fermentation product of waste organic sources. It is found as a component of the effluent or digestate, of anaerobic digestion (AD) processes such as algal AD. Propionate was proposed as an attractive option for a sustainable source of electron donor. An LFCR fed with synthetic propionate showed sulphate reduction occurring via the utilisation of both propionate as well as acetate produced from propionate metabolism. Fermentative bacteria were seen to work syntrophically within the system availing a significant amount of acetate to the SRB community. This acetate was the preferred electron donor over propionate. Maximum volumetric sulphate reduction rates (VSRRs) of 190 mg/L/day were achieved in the reactor. However, detailed analysis showed few propionate-utilising SRB in the community. It was concluded that a more diverse inoculum was needed to investigate the potential of propionate more fully. Secondly, the study investigated lactate as a means to explore the efficiency of electron donors that are incompletely oxidised to acetate. Higher chain VFAs such as lactate are partially oxidised to acetate under biosulphidogenesis; however, acetate oxidation by SRB appears to be a rate-limiting step in most systems. Simultaneous incomplete oxidation of the more complex VFAs with complete oxidation of acetate are rarely reported. Acetate accumulates in the effluents of these processes and results in high chemical oxygen demand (COD) remaining which can lead to environmental impacts such as eutrophication if released into river systems. Further, utilisation of the electron donor is inefficient. In order to address this, the study presents a sequential LFCR system to increase utilisation efficiency of the incompletely oxidised VFA feed. The sequential system was developed by coupling a second reactor unit, specifically colonised with acetate-utilising SRB, to a primary reactor unit utilising lactate. The SRB community in the secondary reactor was able to oxidise acetate from the primary reactor resulting in further sulphate reduction and lower residual COD levels. Residual acetate decreased from 475 mg/L in the primary reactor to 275 mg/L in the secondary reactor. Similarly, residual sulphate decreased from 533 mg/L in the primary reactor to 150 mg/L in the secondary reactor, on a 1 g/L sulphate feed, achieving a much improved effluent water quality. A VSRR of 213 mg/L/day across the sequential system, an overall conversion of 85% and a two-fold increase in sulphate reduced / lactate consumed from 0.45 to 0.85 (g/g) were achieved. Lastly, this study investigated sulphide removal via the incorporation of elemental sulphur into a floating sulphur biofilm (FSB). The generation of an FSB within a sulphate reducing bioreactor is a passive and sustainable method of sulphide remediation, producing a value-added product of elemental sulphur. The sequential reactor system resulted in an increase of two to three-fold in the amount of elemental sulphur recovered, with an improved conversion of sulphide formed to elemental sulphur. Further, the effect of the feed inorganic ions, magnesium and phosphate, on sulphur yield and sulphide removal was studied using the sequential LFCR system. It was found that a decrease of magnesium in the media supplied resulted in an increase in sulphide conversion to sulphur from 21 to 39%, while a concomitant reduction in feed phosphate resulted in a further increase to 50% of the same. In both cases the sulphur concentration in the FSB was substantially increased. Overall, the thesis addresses the need to decrease the cost associated with the supply of a suitable electron donor, as well as improves the water quality of a BSR effluent both in terms of residual sulphate, sulphide and COD. The LFCR system studied was improved both by a sequential reactor system, allowing greater sulphate reduction, sulphide removal via elemental sulphur recovery, and substrate utilisation efficiency. Additionally, changes to the inorganic component of the feed led to further sulphide removal and elemental sulphur recovery. Propionate was concluded to have been only partially used as an electron donor for SRB, however fermentative bacteria present in the mixed community degraded propionate to acetate which was then used for BSR. The work presented here contributes towards the broader research into a semi-passive, environmentally sustainable and economically viable treatment solution for low-flow, circumneutral ARD.

Includes bibliographical references. / The mining and beneficiation of coal and hard rock ores generates large volumes of sulphidic waste that may oxidise in the presence of oxygen and result in the generation of acid rock drainage (ARD). In order to effectively manage the long term effects of ARD, there is a need to reliably quantify the associated impacts and how these impacts evolve with time. Traditional laboratory-scale tests only provide a partial picture of ARD generation, and their extrapolation to full-scale deposits is highly uncertain and controversial. This has prompted the development of mathematical models which take into account the governing chemical reaction and physical transport mechanisms. Whilst the accurate and reliable quantification of the time-related ARD profiles requires rigorous mechanistic modeling of both the (bio) chemical reaction and physical transport mechanisms under non-ideal flow conditions, advanced models are complex and only suitable for site-specific studies and operational decision-making contexts. However, in the early stage screening of waste for potential environmental impacts, simple geochemical mass transport models such as PHREEQC can be used. PHREEQC V.2 has capabilities to simulate a wide range of processes that include equilibrium controlled reactions, kinetically controlled reactions and 1-D advective-dispersion transport, and has been used in a wide range of geochemical applications. However, despite its capabilities, little has been published on its applications to ARD prediction. This study focused on the development and application of a PHREEQC based predictive modeling tool, suitable for the early or screening evaluation of the potential long-term ARD risks associated with sulphidic waste deposits.

The cultivation of bacteria requires high levels of agitation and aeration to satisfy the mass transfer requirements of the cells. Associated with these conditions are turbulent forces which may act on the surfaces of the cells and be detrimental to their growth, metabolism and morphology. Reactor design and operation may require a compromise between the stress sensitivity and the mass transfer requirements of the .microbial system. In this project, the development of general predictive techniques to optimise the design and operation for stress sensitive microorganisms was sought. The effect of hydrodynamic stress on the growth, metabolism and morphology of Corynebacterium glutamicum (ATCC 13032) and Brevibacterium flavum (NRRL 114 75) was investigated in a stirred tank bioreactor in the absence of mass transfer limiting conditions. The results showed that hydrodynamic trauma had no effect on the growth and metabolism of the bacteria. Breakup of bacterial aggregates was however observed, the extent of which depended on the intensity of the hydrodynamic conditions. The extent of aggregate breakup was greater for Corynebacterium glutamicum. It was postulated that the bacteria were held together by a growth associated biomolecular adhesive. The kinetics and mechanism for Corynebacterium glutamicum aggregate breakup in the absence of a gaseous phase, were studied in a stirred tank reactor and capillary flowloop system. A model was developed to describe the rate of aggregate breakup caused by aggregate-turbulent eddy interactions in the impeller zone of the stirred tank reactor and in the wall region of the capillary, in the absence of air bubbles. It assumes that aggregate disruption is caused by the interaction with similarly sized eddies. The extent of aggregate breakup was a function of the magnitude of the turbulent force as well as the total duration of the force event. A similar model was developed to describe the rate of microbial cell death. The forces associated with turbulent eddies in the impeller zone of the stirred tank reactor were compared with those of collapsing air bubbles at the air medium interface. The results showed that both contributed to the total force acting on the cells.