A dissertation submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg In fulfillment of the requirements for the degree of Master of Science in Geography and Environmental Studies 31 May 2018 / South Africa will be dependent on coal for power generation for many decades to come, before a complete transition is achieved where more energy will be generated from non-fossil fuel sources. Through case studies of Hendrina and Kendal Power Stations, this study explored how the management of fly ash (FA) waste in South Africa can be improved to minimise its impact on the environment and human health and examined the potential recycling applications that can benefit local communities. The study drew insights from an environmental justice framework to examine the pollution impacts that FA is exposing to the local community. The environmental justice theory is based on the principle that all people have a right to live in an environment that enhances their wellbeing. Empirical evidence obtained from local community’s in-depth interviews revealed that FA is impacting on the health of communities by exposing them to respiratory and other illnesses and it is also affecting their livelihoods which primarily involves farming. A just transition theory was employed to examine potential socio economic opportunities that can be derived from FA recycling to fulfil redistributive measures that can reduce inequality and eradicate poverty in local communities. Some of Eskom’s power stations like Hendrina are nearing the end of their lifespan since their commissioning in the 1960’s and 1970’s. To aid a just transition, ash recycling was found to have the potential to address the socio economic situation of the power station’s employees and the local communities. The study found that local communities generally lack knowledge about coal ash recycling and need to be empowered and supported to partake in ash recycling ventures. The study argues that a shift in the ash recycling regime is needed in order to benefit local communities and facilitate a just transition to a clean energy production. / MT 2018

Two coal utility plants in South Africa selected (one from Sasol and another from Eskom) for this study produce large volumes of fly ash (over 40 Mt from Eskom at Tutuka, and 3 Mt from Sasol Synfuels at Secunda annually), and brines as by-products during coal processing. Co-disposal of the brines and fly ashes has been a normal practice in these coal-utility plants for decades. Long-term management of fly ash is necessary and requires an understanding and knowledge of how the different waste materials interact with water and brines in different chemical situations. However the geochemistry of their interactions, the leaching and mobility of elements in these disposal systems has not been fully understood. This work gives insights into the chemical processes taking place in the brine-water/brines systems that govern the concentrations of major and minor elements in ash leachates under different environmental conditions. The possible presence of organic compounds (subsequently referred to as 'organics') in brines and their effects on the leaching chemistry of fly ash was also studied. Sustainability and long term impact of the co-disposal of fly ash and brines on the environment was studied through static (batch tests) modeling of the pH-dependent acid neutralization capacity (ANC) tests and columns modeling for dynamic leach tests. The modeling was based on experimental results from other Sasol-Eskom ashbrine project collaborators. Modeling results of the ANC tests were in good agreement with the reported experimental results, which revealed that the release trends of various elements (including trace, heavy elements and contaminants) contained in fly ash into solution is highly pH dependent. However Na, K, Mo and Li exhibited constant solubilisation which was independent of pH changes from all the scenarios. The presence of different constituents of brines subjected to ANC resulted to different ANC capacities ranging from 0.98 moles H⁺/Kg dry ash (of ash-organics mixed with Mg-brines) to 3.87 H⁺/Kg dry ash for those with the C(4) brines. As expected, those constituents from the cationic brines were found on the lower region of acid addition (in the order Mg-brines < Ca-brines < Na-brines) while the anionic brines were found at the upper region of acid addition (in the order S(6)-brines < Cl-brines < C(4)-brines). In the middle region of acid addition were three important scenarios: that of ash with brine, ash without brines (i.e. ash with DMW) and ash with both ASW organics and combined brines. It was from these three scenarios that a generalization of the effect of brines and organics on the ANC was inferred. The ANC of ash with demineralised water (DMW) was 2.33 mol H⁺/Kg dry ash and that of ash with ASW organics lower at 2.12 mol H⁺/Kg dry ash which was the same value as that of ash with combined brines. This indicated that brines decreased the ANC of ash by about 9.01 % and which could be attributed to the acid-base neutralization process and the dynamics of solid phase dissolutions in response to the acid addition. Both fly ashes exhibited a typical pH > 12 (suspension in demineralised water) and the predominant cation even at this high pH is Ca²⁺ (at concentration > 0.002 mmol/L). This indicates that dissolution of CaO and formation of OH⁻ species at pH > 10 contributes to acid neutralisation capacity of both fly ashes and is the greatest contributor to the acid neutralizing capacity of both fly ashes. Two broad leaching behaviours as a function of pH were observed from the three fly ash-ASW organics-brines scenarios (i) leaching of Ca, Mg, Ni and Sr follows a cationic pattern where the concentration decreases monotonically as pH increases; (ii) leaching of Al, Fe, Ti and Zn follow an amphoteric pattern where the concentration increases at acidic and alkaline pH, although Al showed some anomaly from pH 11 where the concentration decreased with the increase in pH. Al showed an amphoteric pattern in which its release increased between pH 12.8 and 11 for all the scenarios and then decreased with decrease in pH down to neutral pH of 7. The batch leaching simulation results from hydrogeochemical modeling also showed that mineral dissolution, precipitation and new phase formation during ash-organics-brines interactions was controlled by pH. The newly formed phases however remain in equilibrium with the ash-brines-organics mixture. Each individual mineral phase dissolution/precipitation/formation system controls the concentration and speciation of the respective constituent elements as evidenced by the log C-pH diagrams obtained from the modeled scenarios. The ash-brines-organics interactions do exhibit and affect the mineralogical chemistry of fly ash. However the extent to which these interactions occur and their effect, varies from one scenario to another, and are dependent on the amounts and type of the constituent brine components. Organics do have a significant effect on dissolution characteristics of few minerals such as calcite, mullite, kaolinite, Ni₂SiO₄, and SrSiO₃ due to complexation effect. The effect is quantitatively conspicuous for calcite mineral phase and for the formation of some new phases such as Fe(OH)₃(am)-CF and portlandite. The composition of the liquid phase from acid neutralisation capacity experiments was successful.Hydrogeochemical modeling was used as a means to provide insights and understanding of the complex reactions taking place, speciation and mineralogical changes occurring. These changes would serve to predict future environmental scenarios when pH conditions change. In this study, an extension of the application field of PHREEQC hydrogeochemical code for modeling and simulation of equilibrium; kinetic and transport mechanisms associated with the interaction of water; and organics and brines with fly ash during their co-disposal is successfully demonstrated. The parameters associated with these mechanisms were used as inputs into the PHREEQC program using modified Lawrence Livermore National Laboratory (LLNL) database for inorganic brines and MINTEQ.V4 database for organics, and used to model the results of ANC test data for the fly ashes. A special reference is made to two separate modeled mineralogical ash recipes from two of the South African power utility plants' fly ash systems, namely, Tutuka and Secunda. The effects of brines in the leaching of major, minor and trace elements at various pH values and the mineralogical changes associated with the intermediate and final products from the interactions of ash-brines systems under different scenarios are qualitatively and quantatively discussed. Multiphase saturation characteristics have been determined for mineral species in contact with water and brines. The modeling results indicated that several mineral phases could be controlling the species concentration in the leachates, and the ANC and column modeling results corroborated well in many aspects with the experimental results obtained from collaborating institutions (South Africa Universities and Research institutions). In addition, application of the PHREEQC model to the ash heap under different disposal systems was carried out to predict the heap leachate composition and geochemical transformations taking place in a period of time. Pore water chemical analysis, and moisture content analysis revealed that contact of the ash with water is a crucial factor in the mobilization of the contaminants with time. Maximum weathering/dissolution of the ash is observed in the top layer (1-3) m and at the point of contact with the subsurface water level which was in good agreement with the model results. The surface layer and the very lowest layers of the dump in contact with lateral flows experience the highest degree of weathering leading to depletion of species. The geophysical transformation of fly ash was also captured through the porosity change calculations and the results revealed that geochemical reactions do affect the porosity of fly ash during the weathering processes. These modelling results were in agreement with the hydraulic tests and salt leaching tests conducted during Sasol-Eskom ashbrine project in Phase I which suggested that salts captured in the ash will become mobile and leach from the fly ash over time. The data therefore indicates that ash dumps may not act as sustainable salt sinks. These findings may have some bearing on engineering decisions on fly ash reuse. From the above observations, it is apparent that release of large quantities of the salts in the ash depends on the extent of its interaction with brines being used for irrigation or with water, either through plug-in flow after a rainfall event or contact with groundwater. The results revealed effects of brine-water contact time with fly ash, the flow volume and velocity, the pH, the degree of saturation, hydrogeology and ash heap geometry as important factors that affect fly ash transformation and weathering. Overall, the ash heap modeling enhanced the understanding of the ash-brines interactions and demonstrated that leachate composition is determined by the following factors; (i) the mass flows from the pores of fly ash, (ii) the surface dissolution of the mineral phases, (iii) the various chemical reactions involved during the ash-brine and ash-water interactions, (iv) the interactions with a gas phase (atmospheric CO₂), (v) the composition of the initial fly ash, and (vi) by the leachate flow and hydrodynamics as captured in the conceptual model. Any ash handling system should therefore be designed to take these criteria into consideration to prevent environmental contamination. The modeling results also gave indications that the ash-brine co-disposal in dry ash systems would be an unsustainable way of locking up brine salts in the long run. In this Thesis, modeling results were used to support experimental data which further reaffirmed the important role hydrogeochemical modeling plays in liquid and solid waste management. Furthermore, hydrogeochemical modeling complements the work of analytical/environmental scientists as well as guiding the future solid waste management and engineering decisions. / Thesis (Ph.D.)-University of KwaZulu-Natal, Westville, 2012.

MENVSC / Department of Ecology and Resource Management

Dissertation submitted in fulfilment of requirements for the degree Master of Technology: Chemical Engineering In the FACULTY OF ENGINEERING 2014 / In South Africa only 5% of the coal fly ash produced annually by power stations finds use. Due to the high quantities of Si and Al in the coal fly ash researchers have explored the opportunity to use the fly ash as a feedstock in zeolite synthesis. Two principal methods have been successfully employed on a micro scale namely the 2-step method and fusion assisted method. However, in order to scale-up these processes some fundamental process design changes are required. Fly ash contains various elements including highly toxic elements such as As, Pb and Hg. The fate of these elements during the synthesis processes is not known. Both these processes generate large quantities of liquid supernatant waste. Disposal of these wastes would be expensive and environmentally harmful, thus making these processes industrially unfeasible. The well known fusion assisted process, contains an energy intensive fusion step operating at 550 C. Construction and operation of a furnace to implement fusion would be too expensive on an industrial scale. The 2-step method has a time consuming pre-hydrothermal treatment step (aging step). In order to improve the feasibility of the 2-step process the processing time of the aging step needs to be reduced. In order to breach the scale gap between micro and pilot plant scale a principal reactor design has been suggested. However, to date, no consideration has been given to the safety and operational reliability of this design. A HAZOP study is required to prevent costly incidents from occurring during the operation of this reactor. The aim of this study formed part of the overall initiative to scale-up the synthesis of zeolites to pilot and ultimately do at industrial scale. The aim of this study specifically was to perform some principal process design activities in order to prepare these processes for scale-up. The objectives were to perform material balances on the two principals synthesis approaches in order to determine the distributional fate of elements. Secondly, to make critical process design changes and develop protocols whereby the supernatant waste resulting from these processes can be minimised. Thirdly, to replace the fusion step (used in the fusion assisted process) and the aging step (used in the 2-step process) with a short high intensity sonochemical treatment step. Lastly, to perform a HAZOP study on the principal bench scale reactor design, and make design changes based on the outcome of the study. Material balances illustrated that most of the elements originating from the coal fly ash (Fe, Mn, Mg, Ca, Ti, Ba, Ce, Co, Cu, Nb, Ni, Pb, Rb, Sr, Y and Zn) do not leach out into solution during either of the two synthesis approaches. This was due to the CaO content in the ash retarding the mobility of these elements. This meant that during the 2-step process these elements reported to the overall zeolite product but did not form part of the zeolite crystal structure. On the other hand, during the fusion assisted process these elements reported to the solid residue waste. The yield efficiency of the fusion assisted process was found to be poor with only 19.6% of the Si and 21.6% Al reporting to the zeolite A product. The 2-step process on the other hand incorporated 72.2% of the Si and 81.5% Al into the zeolite product. However, the 2-step process produced a mixed phase zeolite product while the fusion assisted process produced a pure phase zeolite A product. Therefore there is a trade-off between yield efficiency and product purity. It was found that the liquid supernatant waste produced during both the synthesis processes contained toxic elements such as As, Pb, Hg, Al and Nb. This highlighted the importance to minimise the liquid supernatant waste generated. The waste minimisation studies illustrated that the liquid supernatant waste can be recycled while still producing highly crystalline zeolite products, in both the synthesis approaches. During the 2-step process the supernatant waste was recycled as a source of NaOH. By recycling the waste it was found that 40% of the supernatant could be recycled. However, by making a minor process design change a protocol was developed whereby 100% of the supernatant waste could be recycled. Also, by recycling the liquid waste, zeolite analcime became the dominant phase due to the accumulation of Si in the waste. In the fusion assisted process, protocols were developed whereby the liquid supernatant waste was recycled as a source of water. It was found that 100% of the supernatant could be recycled without compromising the relative crystallinity and purity of the zeolite A product. Both the fusion step (used in the fusion assisted approach) and the 48 hr aging step (used in the 2-step process) could be replaced with 10 min of sonochemical treatment. It was found in both cases that the introduction of ultrasound, during the pre-hydrothermal stage, increased the rate of crystal formation during the hydrothermal treatment step. It was also found that by replacing the high temperature fusion step, in the fusion assisted process, the required hydrothermal treatment temperature could be reduced to 90 C. By introducing sonochemical treatment in these two synthesis approaches their synthesis time and energy demands could be reduced successfully. A HAZOP study on the principal bench scale reactor design enabled design changes to be made preventing future loss during operation. A final optimised reactor design was proposed based on the outcome of the HAZOP study. This study effectively prepared both zeolite synthesis approaches for up-scale operation. Scale-up of this process will reduce disposal of coal fly ash offering relief to the financial and environmental strain caused to the country.