Mine wastewaters generated during active production operations, and decanting streams following mine closure have major environmental impacts, and volumes requiring treatment are expected to increase substantially as the South African mining industry matures. Biological treatment of mine waters has been the subject of increasing interest, where sulphate reducing bacteria are employed for the reduction of sulphate to sulphide, precipitation of metals and the production of alkalinity. However, the sulphide if not removed from the system can be oxidised back to sulphate. As a result there have been limitations especially in the provision of technological options that are sustainable over the long-term, where the total sulphur (in its different forms) can be removed from the system. These, however, are the subject of a number of constraints including, importantly, the process capability to remove reduced sulphur from the treated stream, in one of its oxidation states, and thus linearise the biological sulphur cycle. This remains a major bottleneck in the development of biological wastewater treatment technology. Floating sulphur biofilms are observed as surface layers in numerous aquatic sulphide-rich environments, and it has been suggested that they play a role in the biological cycling of sulphur. The use of sulphur biofilms for the removal of elemental sulphur was identified in this study as a possible means for addressing the technological bottleneck, especially in passive wastewater treatment systems. There is, however, little documented information in the literature on the structure of floating sulphur biofilms, the microbial species responsible for their occurrence or bio-process applications of the system. A linear flow channel reactor was developed to simulate natural conditions and enabled the study of floating sulphur biofilm under controlled laboratory conditions. It was observed that these biofilms developed through three distinct stages termed Thin, Sticky and Brittle films. A microprobe study showed the presence of a steep Redox gradient established across (260 to 380 μm) depth of the floating sulphur biofilm of ~ 0 to -200 mV (top to bottom), which correlated with pH and sulphide gradients across the system. Structural investigations embedded in an exopolymeric matrix containing clearly defined channels and pores. Sulphur crystals were found to develop within the biofilm and above a certain size these disengaged and then settled in the liquid phase below the biofilm. These features, together with the ability of the biofilm to remain suspended at the air/water interface thus provide the surface requirement, and indicate that these structures may be understood as “true” biofilms. In order to study an apparent functional differentiation within the floating sulphur biofilm system, a method was developed to expand its various components over a 13 cm length of agarose tube and across which an oxygen/sulphide gradient was established. This was done by inserting a sulphide plug in the bottom of the tube, overlaying this with the biofilm mixed and suspended in agarose and leaving the tube to open air. After allowing for growth, the different components of the microbial population occurring at various levels across the oxygen/sulphide gradient were sampled. The microbial population was found to resort in distinct functional layers. Aerobes including Acidithiobacillus and Azoarcus, Acidithiobacillus, Thiothrix, Thiovirga and Sulfurimonas were found in the upper oxidised layer. Aerobe and facultative anaerobes such as Chryseobacterium, Bacteroides and Planococcus were found in the middle and heterotrophic anaerobes such as Brevundimonas and uncultured anaerobes were found in the bottom anoxic layer. This enabled the development of a first descriptive structural/functional model accounting for the performance of floating sulphur biofilms. The potential of the floating sulphur biofilm for use as a bioprocess unit operation for sulphide removal in lignocellulose-based low-flow passive systems for acid mine drainage wastewater treatment was investigated. The linear flow channel reactor was scaled up and it was shown that the optimum sulphide removal of 74 % and sulphur recovery of 60 % could be achieved at 20 °C. In a further scale up of the linear channel reactor, the floating sulphur biofilm reactor was developed and operated. Sulphide removal and sulphur recovery of 65 and 56 % respectively was measured in the process. An understanding of the nature and function of floating sulphur biofilms and the further development of their potential application in sulphide removal in aquatic systems may provide a useful contribution to the treatment of acid mine drainage and other sulphidic wastewaters.

An investigation of the biodegradation of complex hydrocarbon wastes, with emphasis on chlorinated aromatic compounds, in an anaerobic/aerobic bioprocess environment was made. A reactor configuration was developed consisting of linked anaerobic and aerobic reactors which served as the model for a proposed bioremediation strategy targeting subterranean soil/sediment/aquifer chlorinated phenol-contaminated environments. Here oxygen is frequently limited and sulphate is readily available, as occurs especially in marine sediment and intertidal habitats. In the anaerobic system the successful transformation and mobilization of the model contaminant, 2,4,6-trichlorophenol, was shown to rely on reductive dechlorination by a sulphate-reducing dependent dechlororespiring co-culture. This was followed in the aerobic system by degradation of the pollutant and its metabolites, 2,4-dichlorophenol, 4-chlorophenol and phenol, by immobilized white-rot fungi.The strategy was initially investigated separately in laboratory bench- and intermediate scale reactors whereafter reactors were linked to simulate the integrated biodegradation strategy. The application of the fungal reactor to treat an actual waste stream by degrading complex mixtures of hydrocarbons in a waste oil recycling effluent was also investigated. The mineralization of phenol and 2,4,6-TCP by immobilized fungal cultures was studied in pinewood chip and foam glass bead-packed trickling reactors. The reactors were operated in sequencing batch format. Removal efficiency increased over time and elevated influent phenol and TCP (800 and 85 mg.L⁻¹) concentrations were degraded by > 98 % in 24 – 30 h batch cycles. Comparable performance between the packing materials was shown. Uptake by the packing was negligible and stripping of compounds induced by aeration had a minimal effect on biodegradation efficiency. Reactor performances are discussed in relation to sequencing batch operation and nutrient requirements necessary to sustain fungal activity in inert vs. organic material packed systems. It was shown that a co-culture consisting of sulphate-reducing and dechlororespiring bacteria established in fed-batch and soil flasks, as well as pine chip-packed fluidized bed reactors. Results showed reductive dechlorination of 2,4,6-TCP to be in strict dependence on the activity of the sulphate-reducing population, sulphate and lactate concentrations. Transformation to 2,4-DCP, 4-CP and phenol was enhanced in sulphate deficient conditions. Dechlororespiring activity was found to be dependent on the fermentative activity of sulphate-reducing bacteria, and the culture was also shown to mobilize and dechlorinate TCP in soils contaminated with the pollutant. Linking the systems achieved degradation of the compound by > 99 % through fungal mineralization of metabolites produced in the dechlororespiring stage of the system. pH correction to the anaerobic reactor was found to be necessary since acidic effluent from the fungal reactor inhibited sulphate reduction and dechlorination. The fungal reactor system was evaluated at intermediate-scale using a complex waste oil recycling effluent. Substantial COD reduction (> 96 % in 48 h batch cycles) and removal of specific effluent hydrocarbon components was shown in diluted, undiluted (COD > 37 g.L⁻¹) and 2,4,6-TCP-spiked effluents. Industrial application of the fungal reactor was evaluated in a 14 m³ pilot plant operated on-site at a waste oil processing plant.

Water is an essential element in all aspects of life and is vital for both domestic and industrial purposes regarding both the quality and quantity thereof. Similar to many other drought stricken countries, South Africa requires water for the socio-economic growth of the country, yet is faced with the problem of maintaining the quality of its drinking water as well as protecting the dwindling supplies. In an attempt to prevent the deterioration of South African water supplies the treatment, purification and recycling of industrial and mining waste water has recently become of prime importance. Many industrial and mining waste waters contain heavy metals in toxic quantities. The conventional processes that have been used till recently to address this problem, are often expensive or contain chemical agents which compound the environmental problem. As an alternative biological methods of metal accumulation appear to offer an economic and efficient alternative to these methods. An advantage to the South African scenario is the commercial production of the yeast, S. cerevisiae as a readily inexpensive by-product from some fermentation industries, Yeast cells, and in particular S. cerevisiae have proven to be capable of accumulating heavy metals, and therefore exhibit potential application in the bioremediation of waste water. The aim of this project was twofold. The initial part of this work attempted to define the mechanisms of metal accumulation by the yeast cells and cellular components. The information obtained from these initial studies provided a data base required for the development of a bioremediation system. Initial contact with the metal ions occurs at the wall interface of the yeast cell. Metal accumulation appears to be a function of all the cell wall components. The isolated cell wall components are better metal chelators then the intact cell walls. An apparent affinity series of mannan > chitin> glucan > intact cell walls exists. However, these components differ in their affinities for metal ions. Storage of metal ions within the cell occurs predominantly in the vacuole. The present study concluded that metal accumulation by the vacuole could be related to size. Metal accumulation occurred in the order of Cu2+ > Co2+ > Cd2+ with a corresponding decrease in atomic radii of Cd2+ > C02+ > Cu2+. Vacuolar ion deposition occurs at an early stage during the internalization of metal ions within the yeast cells. At the onset of vacuolar saturation, depositions of metal ions as granules within the cytosol occurs. In the presence of heavy metal cations viable yeast cells can be shown to exhibit two types of cellular responses. Uptake of Cu2+ and Cd2+ causes the loss of intracellular physiological cations from within the yeast cell. In comparison, uptake of Co2+ into the cell does not have this effect. All three heavy metal cations initiate plasma cell membrane permeability, thus the Cu2+ and Cd2+ induced loss of the intracellular cations, occurs. ~ a result of ion-exchange mechanisms and not due to cation leakage brought about by membrane permeabilization. Uptake of heavy metals by viable yeasts appears to be generally non-selective though the amount of metals accumulated are largely affected by the ratio of ambient metal concentration to biomass quantity. In addition, the energy dependent nature of internalization necessitates the availability of an external energy source for metal uptake by viable yeast cells. For these reasons metal removal from industrial waste water was investigated using non-viable biomass. By immobilizing the yeast cells additional mechanical integrity and stability was conferred apon the biomass. The three types of biomass preparations developed in this study, viz. polyvinyl alcohol (PV A) Na-alginate, PV A Na-orthophosphate and alkali treated polyethylenimine (PEI):glutaraldehyde (GA) biomass pellets, all fulfilled the necessary physical requirements. However, the superior metal accumulating properties of the PEI:GA biomass determined its selection as a biosorbent for bioremediation purposes. Biosorption of heavy metals by PEI:GA biomass is of a competitive nature, with the amount of metal accumulated influenced by the availability of the metal ions. This availability is largely determined by the solution pH. At low pH values the affinity of the biomass for metals decreases, whilst enhanced metal biosorption occurs at higher pHs, ego pH 4.5 - 6.0. PEI:GA biomass pellets can be implemented -as a biosorbent for the bi9remediaiton of high concentration, low-volume metal containing industrial waste. Several options regarding the bioremediation system are available. Depending on the concentration of the metals in the effluent, the bioremediation process can either be used independently or as part of a biphasic remediation system for the treatment of waste water. Initial phase chemical modification may be required, whilst two types of biological systems can be implemented as 'part of the second phase. The PEI:GA biomass can either be contained within continuous-flow fixed bed tanks or continuous-flow stirred bioreactor tanks. Due to the simplicity of the process and the ease with which scale-up is facilitated, the second type of system shows greater application potential for the treatment of this type of industrial waste water than the fixed-bed systems.

The hydrolysis of complex organic heteropolymers contained in municipal wastewater to simpler monomers by extracellular hydrolytic enzymes is generally considered the rate-limiting step of the biodegradation process. Previous studies of the Recycling Sludge Bed Reactor (RSBR) revealed that the hydrolysis of complex particulate organics, such as those contained in primary sludge (PS), was enhanced under anaerobic biosulphidogenic conditions. Although the mechanism was not fully understood, it appeared to involve the interaction of sulfide and sludge flocs. The current study was conducted using a 3500 ml laboratory-scale RSBR fed sieved PS at a loading rate of 0.5 kg COD/m³.day and an initial chemical oxygen demand (COD) to sulfate ratio (COD:SO₄) of 1:1. There was no significant accumulation of undigested sludge in the reactor over the 60-day experimental period and the quantity of SO₄ reduced indicated that the yield of soluble products from PS was at least as high as those reported previously for this system (> 50%). In the current study, the specific activities of a range of extracellular hydrolytic enzymes (L-alanine aminopeptidase, L-leucine aminopeptidase, arylsulphatase, α-glucosidase, β- glucosidase, protease and lipase) were monitored in a sulfide gradient within a biosulphidogenic RSBR. Data obtained indicated that the specific enzymatic activities increased with the depth of the RSBR and also correlated with a number of the physicochemical parameters including sulfide, alkalinity and sulfate. The activities of α- glucosidase and β-glucosidase were higher than that of the other enzymes studied. Lipase activity was relatively low and studies conducted on the enzyme-enzyme interaction using specific enzyme inhibitors indicated that lipases were probably being digested by the proteases. Further studies to determine the impact of sulfide on the enzymes, showed an increase in the enzyme activity with increasing sulfide concentration. Possible direct affects were investigated by looking for changes in the Michaelis constant (Km) and the maximal velocity (Vmax) of the crude enzymes with varying sulfide concentrations (250, 400 and 500 mg/l) using natural and synthetic substrates. The results showed no significant difference in both the Km and the Vmax for any of the hydrolytic enzymes except for the protease. The latter showed a statistically significant increase in the Km with increasing sulfide concentration. Although this indicated a direct interaction, this difference was not large enough to be of biochemical significance and was consequently not solely responsible for the enhanced hydrolysis observed in the RSBR. Investigation into the floc characteristics indicated that the biosulphidogenic RSBR flocs were generally small in size and became more dendritic with the depth of the RSBR. Based on the above data, the previously proposed descriptive models of enhanced hydrolysis of particulate organic matter in a biosulphidogenic RSBR has been revised. It is thought that the effect of sulfide on the hydrolysis step is primarily indirect and that the reduction in floc size and alteration of the floc shape to a more dendritic form is central to the success of the process.

Brewery effluent that had undergone treatment in an anaerobic digester (AD) was used as an alternative water and nutrient source for hydroponic crop production. Brewery effluent was demonstrated to contain sufficient nutrients to support the growth, flowering and fruiting of Lycopersicum escolentum "Moneymaker" tomato crops. The adjustment of the effluent pH with phosphoric acid to between pH 6.0 and 6.5 increased the development of the crops by around 100% compared to crops grown in unaltered effluent. The pH adjusted effluent-grown plants grew to a mean height of 831.4 ± 21.1 mm and a dry biomass weight of 42.34 ± 2.76 g compared to the unaltered pH effluent plants which grew to a height of 410.6 ± 20.5 mm and a weight of 7.65 ± 0.68 g after 49 days. Effluent treatment in high-rate algal ponds (HRAP) was determined to have no positive effect on the nutritional potential of the effluent for Moneymaker production. The effluent-grown plants did not perform as well as plants grown in inorganic-fertilizer and municipal water. Plants grown in effluent grew taller but did not produce significantly more fruit when phosphoric acid (height: 1573.3 ± 50.4 mm, 19.4 ± 1.4 fruit per plant) was compared to nitric acid (height: 1254.1 ± 25.4 mm, 15.6 ± 1.5 fruit per plant) as the pH adjustment over 72 days. Direct and secondary plant stresses from effluent alkalinity, ammonium nutrition, nitrogen limitation, sodium concentrations and heat stress among other factors were probably confounding variables in these trials and require further investigation. Considering the raw effluent composition and manipulating the AD operation is a potential opportunity to improve overall AD performance, reduce chemical inputs in the effluent treatment process, reduce the final effluent alkalinity, and increase available nitrogen content in the final effluent. The anaerobic digester discharging >1000 m³ of nutrient enriched effluent every day is a resource with considerable potential. The benefits of developing this resource can contribute to cost-reduction at the brewery, more efficient water, nutrient and energy management at the brewery, and offer opportunities for job creation and potentially benefit local food security.

Maintenance of membrane performance requires inevitable cleaning or defouling of fouled membranes. Membrane cleaning using enzymes such as proteases, lipases, α-glucosidases from a sulphidogenic bioreactor was investigated. At first, dilute and concentrated enzyme extract were prepared form the sulphidogenic pellet. Enzyme assays on 0.5 % azocaisen, 1 % triacetin and 1 mg/ml ρ-nitrophenyl-α-D-glucopyranoside were performed using the concentrated enzyme extract (0 – 200 mg/ml). For membrane fouling, an abattoir effluent was obtained from Ostritech Pty (Ltd), Grahamstown, South Africa. The effluent was characterised for presence of potential foulants such as lipids, proteins, amino acids and carbohydrates. Static fouling of polysulphone membranes (0.22 μm, 47 mm) was then performed using the abattoir effluent. Cleaning of the fouled membranes was also performed using at first the dilute and then the concentrated form (200 mg/ml) of enzyme extracts. Qualitative and quantitative biochemical analysis for proteins, lipids and carbohydrates was performed to ascertain the presence of foulants on polysulphone membranes and their removal by dilute or concentrated enzyme extracts. The ability of dilute enzyme extracts to remove proteins lipids, and carbohydrates fouling capillary UF membrane module; their ability to restore permeate fluxes and transmembrane pressure after cleaning/defouling was also investigated. Permeate volumes from this UF membrane module were analysed for protein, amino acids, lipids, and carbohydrates concentrations after fouling and defouling. Fouling was further characterized by standard blocking, cake filtration and pore blocking models using stirred UF cell and polyethersulphone membranes with MWCO of 30 000, 100 000 and 300 000. After characterization of fouling, polyethersulphone membranes with MWCO of 30 000 and 300 000 were defouled using the concentrated enzyme extract (100 mg ml). Enzyme activities at 200 mg/ml of enzyme concentration were 8.071 IU, 86.71 IU and 789.02 IU for proteases, lipases and α-glucosidases. The abattoir effluent contained 553 μg/ml of lipid, 301 μg/ml of protein, 141 μg/ml of total carbohydrate, and 0.63 μg/ml of total reducing sugars. Proteins, lipids and carbohydrates fouling polysulphone membranes after a day were removed by 23.4 %, when a dilute enzyme was used. A concentrated enzyme extract of 200 mg/ml was able to remove proteins, lipids and carbohydrates up to 5 days of fouling by 100 %, 82 %, 71 %, 68 % and 76 % respectively. Defouling of dynamically fouled capillary ultrafiltration membranes using sulphidogenic proteases was successful at pH 10, 37°C, within 1 hour. Sulphidogenic proteases activity was 2.1 U/ml and flux Recovery (FR %) was 64. Characterization of fouling revealed that proteins and lipids were major foulants while low concentration of carbohydrates fouled polyethersulphone membranes. Fouling followed standard blocking for 10 minutes in all the membranes; afterwards fouling adopted cake filtration model for membranes with 30 000 MWCO and pore blocking model for membranes with 300 000 MWCO. A concentration of 100 mg/ml of enzyme extract was able to remove fouling from membranes with MWCO of 30 000. Defouling membranes that followed pore blocking model i.e. 300 000 MWCO was not successful due to a mass transfer problem. From the results of defouling of 30 000 and 300 000 MWCO it was concluded that defouling of cake layer fouling (30 000 MWCO) was successful while defouling of pore blocking fouling was unsuccessful due to a mass transfer problem. The ratio of enzymes present in the enzyme extract when calculated based on enzymatic activity for proteases, lipases and α-glucosidases was 1.1 %, 11 % and 87.9 %. It was hypothesized that apart from proteases, lipases, α and β-glucosidases; phosphatases, sulphatases, amonipeptidases etc. from a sulphidogenic bioreactor clean or defoul cake layer fouling by organic foulants and pore blocking fouling provided the mass transfer problem is solved. However, concentration of enzymes from a sulphidogenic bioreactor has not been optimized yet. Other methods of concentrating the enzyme extract can be investigated for example use of organic solvents.

Please read the abstract in the front section of this document / Dissertation (MSc (Microbiology))--University of Pretoria, 2006. / Microbiology and Plant Pathology / unrestricted

No abstract available. / Dissertation (MSc (Water Utilization))--University of Pretoria, 2006. / Chemical Engineering / unrestricted

Please read the abstract in the section 00front of this document. / Thesis (PhD (Chemistry))--University of Pretoria, 2008. / Chemistry / unrestricted

The activated sludge process comprises a complex and enriched culture of a mixture of generalist and specialist organisms. The lack of knowledge on species diversity of microbial communities is due to the simplicity of bacterial morphology and the phenotypic characters, and the unculturable portion of microbial cells in natural habitats. Although a wide range of bacteria can be isolated using conventional microbiological techniques of sample dilution and spread plate inoculation, many well-known activated sludge bacteria can not be isolated using them. The individual microbial cells in activated sludge grow in aggregates that consist of floc-forming organisms together with filamentous microorganisms that form the backbone of the activated sludge floes. Overgrowth of these filamentous microorganisms often causes settling problems called bulking and foaming. These problems consist of slow settling, poor compaction of solids and foam overflow into the effluent. Although methods for the isolation of filamentous bacteria from mixed liquor samples have been investigated, the attempts have been largely unsuccessful. In this study we investigated bulking and foaming activated sludge to identify the dominant filamentous organisms using microscopy and molecular techniques. Using microscopy, the dominant filament associated with the foaming sample was "Microthrix parvicella" and in the bulking sample was Nocardia spp. The foaming sample was investigated using molecular techniques that involved 165 rDNA sequencing. Although some of the clones isolated from the sludge foam were associated with filamentous bacteria causing foam, no positive identification could be made. In the part of the study that was conducted in Australia, a rRNA-targeted oligonucleotide probe was designed for the identification of a filamentous organism occurring in activated sludge foam. This organism resembled Eikelboom Type 0041 and was classified in the candidate bacterial division TM7. The discrepancy that the sequence data did not indicate the dominant filamentous organisms observed by microscopy, highlights the fact that natural microbial communities need to be studied using a combination of techniques since none of the techniques available are sufficient to determine the complete community structure of complex communities such as activated sludge. / Dissertation (MSc (Microbiology))--University of Pretoria, 2005. / Microbiology and Plant Pathology / unrestricted