School of Chemical and Metallurgical Engineering (ETDs)
Permanent URI for this communityhttps://hdl.handle.net/10539/38875
Browse
8 results
Search Results
Item Co-gasification of Coal and Solid Waste to Hydrogen Enriched-Syngas in a Fixed Bed Gasifier(University of the Witwatersrand, Johannesburg, 2020-10) Ozonoh, Maxwell; Daramola, Michael O.; Oboirien, Bilainu O.The economic growth of every nation around the globe is centred on energy. Energy can be harnessed from different sources using different conversion systems, but such systems should be sustainable. Liquid fuels such as petroleum and solid fuels (e.g. coal & biomass) are largely used for energy production. Energy recovery from these fuels is usually carried out using thermal chemical processes such as combustion, pyrolysis, and gasification systems. Out of the three technologies, gasification is considered the most attractive based on its efficiency and other qualities. In the gasification process, syngas is produced. It is necessary to produce syngas of high quality such as hydrogen-enriched syngas. Hydrogen-enriched syngas can be used in fuel cells, gas turbines and engines for electricity production. This type of gas burns with little gaseous emissions to the atmosphere, but its production is dependent on the type of fuel and process conditions, and energy conversion system employed. In South Africa, around 95 % of electric power production comes from coal, and the current reserve is projected to last not more than a century [8]. Secondly, the coal is fast depleting and generates a lot of gaseous emissions (e.g. CO2, NOX & SOX) that pose a huge threat to the environment. The emission of the aforementioned gases is a very serious issue in South Africa. Presently, some Carbon Capture and Storage (CCS) projects are on-going in the country, although the CCS is not the fuse of this study. The gasification of biomass waste and coal could assist in gaseous emission reduction. Similarly, large amounts of agricultural wastes (e.g. sugarcane bagasse, corn cob & pine saw dust) and other solid waste such as tyre are in abundance in SA. It is detailed in chapter 2. Majority of the wastes are disposed indiscriminately, hence resulting in environmental pollution. Importantly, the solitary gasification of biomass is very expensive considering the prices of biomass. Besides that, biomass produces large amount of tar hence, resulting in operational difficulties in the gasifier and end user facilities. In this study, co-gasification of coal and solid wastes is considered as a crucial alternative to addressing the aforementioned problems. Particularly, the feedstocks used for this study were coal, biomass (corn cob (CC), pine sawdust (PSD), sugarcane bagasse (SCB)) and waste tyre (WT) and were pre-treated by drying, milling, sieving, and torrefaction (coal was not torrefied). The fuel samples were blended with coal at different ratios as detailed in the thesis and used for the study. For the torrefaction process, the most viable torrefaction process conditions and feedstock were determined, while the torrefaction process model for the feedstocks were developed, using Response Surface Methodology (RSM) and Artificial Neural Network (ANN), respectively. The Performance efficiency of gasification systems was evaluated using experimental data obtained from a few gasifiers (e.g. entrained, fluidised, and fixed bed) operated at varied experimental conditions using blends of feedstocks (e.g. biomass, coal, waste tyre etc.). A backpropagation Levenberg Marquardt (L-M) and Bayesian Regularisation (BR) algorithms of ANN model with Multiple Input- Multiple Output (MIMO) and Multiple Input-Single Output (MISO) layer networks were considered. The results of the MIMO and MISO layer networks obtained from the L-M algorithm were better than that of BR algorithm which is in affirmation with some of the results found in the literature. For model result improvement, Input Variables Representation Technique-by-Visual Inspection Method (IVRT-VIM) and Output Variables Representation Technique-by-Visual Inspection Method (OVRT-VIM) were developed from the study. Estimation of the gaseous emissions and profits from biomass, tyre, and coal fired co-gasification CHP Plant using Artificial Neural Network (ANN) was carried out for 20-year investment period using South Africa (SA) and Nigeria as cases studies via Artificial Neural Network (ANN). Higher profits were obtained from South African feedstocks than that of Nigerian feedstocks due to cheaper price of SA coal WFO and WOFC, but the gaseous emissions (CO, NOX, & SO2) from the Nigerian fuels were lower than that of SA because of differences in compositions of the fuels. The potentials of biomass torrefaction in terms of profitability in a co-gasification CHP plant for a 20-year-investment period was carried out using blends of Coal + SCB, Coal + CC, and Coal + PSD with coal-to-biomass ratio of 50:50, 71:29, and 80:20, respectively. The two financial cases mentioned earlier were considered. Four investment terms including: (A) 1st–5th, (B) 5th– 10th, (C) 10th– 15th & (D) 15th– 20th and two operational cost models; with feedstock costing (WFC) and without feedstock costing (WOFC) were employed. An estimated profit of between USD5.9 million - USD6.5 million and USD7.8 - USD7.9 million was earned at the end of investment plan using WFC and WOFC, respectively. The Internal Rate of Return (IRR) was 5 ± 1 %/yr. and 7 ± 4 %/yr. based on South African electricity price of 0.14 $/c kWh, respectively. The parametric effect of process variables during torrefaction of coal/biomass/waste tyre blends using ANN and RSM models were studied. The variables considered were Higher Heating Value (HHV), Enhancement Factor (EF), and Sold Yield (SY). The most effective operating process conditions (in terms of blending ratio, temperature and torrefaction time: input variables) is of the order: 50:50 at 300 OC and 45 min > 50:50 at 250 OC and 30 min >50:50 at 200 OC and 45 min. Similarly, the most viable fuel follows the order of Coal + Torrefied PSD > Coal + Torrefied SCB > Coal + Torrefied CC and > Coal + Torrefied WT. Coal + Torrefied PSD has HHV of 28.27 % and an EF of 1.41. This corresponded to around 10 % increase in the HHV of the torrefied fuel when compared to the raw fuel and about 25.23% higher than the EF of Coal + Torrefied WT of 1.03. Based on the result of the EF of Coal +Torrefied waste tyre, upgrading of the fuel quality via torrefaction is not recommended. Furthermore, a comprehensive study on tar treatment techniques was carried out using tars produced from biomass and blends of biomass and coal employing biochar based and Ni-biochar based catalysts. Box Behnken Design of Experiment (DoE) method was used. A full quadratic regression model was used to develop a mathematical model for tar treatment based on the feedstocks studied. The Pine Sawdust-Biochar Catalyst (PSD-BC) and Nickel Pine Sawdust-Biochar Catalyst (Ni-PSD-BC) were the most effective in terms of tar treatment and with an average percentage amount of tar conversion of 89.76 and 96.73%, respectively. Ni-PSD-BC was more efficient for tar cracking than PSD-BC, but PSD-BC (waste base) may be more attractive if sustainability and cost effectiveness of precursors are considered. Co-gasification of coal and pine sawdust (PSD) to hydrogen enriched syngas in a fixed bed gasifier was carried out with catalyst (WCAT) at 900 OC and without catalyst (WOCAT), at 700, 800, and 900 OC, respectively. Coal-to-PSD ratio of 1:1 was used, while Nickel-pine sawdust-biochar (Ni-PSD-BC) and pine sawdust-biochar (PSD-BC) were employed as catalysts. The gases produced at 700, 800 & 900 OC using WOCAT cannot be used in fuel cells and gas turbines due to poor quality, while others produced at 900 OC WCAT, can be used in internal combustion engines and gas turbines, but unfortunately, have lower quality to be employed in fuel cells for electricity production. However, the study provides a method of beneficiation of the high ash content South African coal for energy production. The outcome of this study is also instrumental to energy security, efficiency and sustainability as well as waste management in South Africa, Nigeria and other parts of the globe. An assessment of the economic, energy and environmental viability of a 5 MW co- gasification power plant was carried out, using blends of coal and biomass, and two financial cases were considered namely: with feedstock costing (WFC) and without feedstock costing (WOFC). Feedstock profitability in the plant for energy production was evaluated. Equipment consisting was not considered. The power plant used 20,473,451.41 kg, 20,986,049.96 kg, 18,251,806.49 kg, and 15,276,277.85 kg of Coal + SCB, Coal + CC, Coal + PSD, and Coal + WT to produce the 5 MW and 5.56 MW electric and thermal power, annually. Coal + Torrefied PSD was the most profitable of the fuels studied. The use of Coal-to-PSD ratio of 4:1 for the power generation as against Coal-to-PSD blend ratio of 1:1 resulted to an annual loss of about ZAR6, 461,301.77 ($90,458,224.70) and ZAR123,782.47 ($1,732954.58) WFC and WOFC, respectively.Item Thermo-mechanical processing and testing of titanium alloys for potential dental applications(University of the Witwatersrand, Johannesburg, 2022-12) Nape, Kgetjepe Tlhologelo; Chown, Lesley H.; Cornish, LesleyNew titanium alloy compositions were identified for potential dental implants on the basis of having two-phase microstructures for good mechanical properties and by avoiding problematic elements to increase biocompatibility. The Thermo-Calc program with the TTTI3 (TT Ti-alloy) database was used to calculate new Ti compositions, without toxic Al and V as alloying elements. The aim was to mimic the α+β phase proportions in Ti-6Al-4V and Ti-10.1Ta-1.7Nb-1.6Zr (TTNZ) (an analogue for Ti-6Al-4V). Copper (Cu = 1, 3, 5 and 10 wt%) was varied to give the Ti2Cu phase, which gives good hardness and antibacterial properties. A cost analysis was done and the less expensive Ti-6Nb-4Zr-xCu and Ti-8Nb-4Zr-xCu (x = 0 and 5 wt%) compositions were selected for experimental work. The samples were made by arc-melting and prepared for microstructural studies to understand the influence of alloying elements, and to compare with the commercial Ti-6Al-4V and reported Ti-10.1Ta-1.7Nb-1.6Zr (TTNZ) alloys. Hot deformation of the as-received Ti-6Al-4V and TTNZ alloys was investigated, using a Gleeble 3500® Thermo-mechanical Simulation Facility, at 850°C and 950°C and strain rates of 0.1 s-1 and 10 s-1. The as-cast Ti-6Nb-4Zr-xCu and Ti-8Nb-4Zr-xCu (x = 0 and 5 wt%) alloys comprised αTi and βTi, with Ti2Cu once Cu was added, although EDX indicated some inhomogeneity. The XRD analyses identified αTi and small amounts of βTi with solid solution (shifted peaks), with some Ti2Cu. The Ti-8Nb-4Zr alloy (285 ± 7 HV) had similar hardness to Ti-6Nb-4Zr (280 ± 13 HV), and was considered the better alloy. Adding 5 wt% Cu increased the hardness due to Ti2Cu. With the Gleeble, deformation at 950°C and 10 s-1 led to a finer Ti-6Al-4V microstructure, whereas finer Ti-10.1Ta-1.7Nb-1.6Zr (TTNZ) microstructures occurred at 850°C and 10 s-1. The XRD of all deformed Ti-6Al-4V and Ti-10.1Ta-1.7Nb-1.6Zr samples indicated αTi and βTi, with shifted βTi peaks. The Ti-6Al-4V (324 ± 9 HV) deformed at 850°C and 0.1 s-1 had higher hardness than both deformed TTNZ samples. Higher flow stress were obtained at higher strain rate (10 s-1) and lower temperature (850°C). The Ti-6Al-4V alloy had higher flow stress than the TTNZ alloy. Therefore, the TTNZ alloy was considered better, due to its lower flow stress, which indicated better formability. The new alloys had similar hardnesses to Ti-6Al-4V, and were higher than for TTNZ, suggesting that they might have similar properties to Ti-6Al-4V.Item Influence of copper on the corrosion and mechanical properties of Grade 4 titanium for biomedical applications(University of the Witwatersrand, Johannesburg, 2022-12) Hadebe, Nomsombuluko Dayanda Elizabeth; Cornish, Lesley; Chown, Lesley H.; Smit, Melanie; Mwamba, AlainThis study assessed the effect of Ti2Cu and its proportions on the corrosion resistance, and compared the results to Grade 4 commercially pure titanium. The Thermo-Calc program with the TTTI3 (Ti-alloy) database was used to predict the phases. Materials Studio software was used to model the crystal structures and XRD patterns of the phases of Ti-Cu alloys. Ti-Cu samples with 0, 5, 15, 25, 33, 40, 47 and 50 wt % Cu were produced. Composition, microstructures, phases, hardness and corrosion resistance were studied in the as-cast and annealed conditions (750° and 900°C water quenched). The CP Ti samples comprised basket-weave acicular microstructures. The Ti-5Cu samples comprised lamellar (αTi) and Ti2Cu phases. The Ti-15Cu, Ti-25Cu and Ti-33Cu alloys comprised (αTi) dendrites and sparse eutectic of Ti2Cu and (αTi). The ((βTi) dendrites decomposed to (αTi) and Ti2Cu, and could not be retained due to insufficient fast quenching. The Ti-40Cu and Ti-47Cu samples had minor titanium oxide dendrites which solidified first and then Ti2Cu nucleated on them and grew as dendrites, surrounded by the Ti2Cu + TiCu eutectic. In the Ti-50Cu sample, TiCu was the true primary phase and grew as needles, and was subsequently surrounded by a coarse TiCu + Ti2Cu eutectic. No Ti3Cu phase was observed. The microstructures of the as-cast alloys agreed with the Cu-Ti phase diagram of Ansara et al. (2021) and Dyal Ukabhai et al. (2022) with the congruent formation of Ti2Cu, as well as no Ti3Cu. The addition of copper to titanium increased the hardness, while annealing decreased the hardness of the Ti-Cu alloys. Addition of copper above 5 wt % Cu and annealing decreased the corrosion resistance of the samples, but since copper ions in liquid solutions promote the antimicrobial activity, some corrosion is necessary to allow the copper ions to be available. The corrosion tests showed that the corrosion rates obtained were very low, below 0.13 mm/y, which is an acceptable corrosion rate for biomaterial applications. Ti-5Cu showed the best corrosion resistance.Item Upgrading Semi-Soft Coking Coal by Hydrothermal Treatment: Caking and structural properties(University of the Witwatersrand, Johannesburg, 2024-02) Ndumo, Jabulile; Bada, SamsonBased on the current challenges faced by the metallurgical industry in South Africa in importing quality reductants, there is an urgent need to investigate a new approach to enhance the semi-soft coal available in the country. Importing prime coking coal has increased the steel price, resulting in many downstream operations involving steel closing down in the country. With a surplus of semi-soft coking coal in South Africa, this research sought to look into this kind of abundant coal to enhance its property as a reductant for blast furnace applications. For this reason, a study was conducted on two Southern African coals, Grootegeluk (semi-soft coking coal) and Moatize (higher quality coal). Both coals were individually hydrothermally treated and then blended at different ratios to further upgrade their metallurgical properties. The as-received Moatize coal showed properties that were more of prime coking coal with high total carbon content (76.50%), a crucible swelling index of nine, a maximum dilatation of 59% and volatile matter of 20.39%. It was a highly vitrinite coal with a vitrinite reflectance of 1.28%, a higher micropore volume than mesopore volume and a very low maximum fluidity of 24 dial divisions per minute (ddpm). According to the initial test, the Grootegeluk coal sample had a crucible swelling index of 5.5, a high volatile matter of 35.02% and a low vitrinite reflectance of 0.72%. In addition, the sample had a maximum dilatation of -10%, a maximum fluidity of 3ddpm and a higher mesopore volume than the Moatize coal. Hydrothermal treatment was conducted on the coal samples at numerous temperatures (100ºC to 200ºC), at various residence times (30 to 90 minutes) and at different coal masses (300 to 600grams (g)). According to the results, the optimal hydrothermal conditions were 200ºC, 90 minutes and 600g. Another hydrothermal treatment was performed at a higher temperature and residence time of 280ºC and 180 minutes. The same sample mass of 600g was used and the result showed no further improvement. The coal samples were then blended at various Grootegeluk/Moatize ratios (15% to 50% Grootegeluk), and further hydrothermal treatment tests were carried out based on the optimum conditions achieved. Both the hydrothermal test and the blending of the coal led to a coal with volatile matter ranging from 21.46% to 23.79%, which is a required specification for metallurgical application. The total carbon of the enhanced coal blend also ranged from 68.8% to 82.10%, with total sulphur below 1%. The mesopore-micropore ratio of the treated blend was higher than the individual coal samples, which is what is expected of a metallurgical coal. Based on these findings, coke was produced and analysed to identify a coke capable of withstanding blast furnace conditions. Using the particle reactivity index (PRI), proximate analysis and the pore size distribution, 90-(50% Grootegeluk+50% Moatize)-C product was identified as the coke with the least PRI and high fixed carbon. Further investigation showed that the blending and hydrothermal treatment affected the coal’s physiochemical, rheological and micro-molecular properties. The study has established that metallurgical properties of the locally mined semi-soft coking can be enhanced solely and when mixed with a hard coal. Even though the 90-(50%GG+50%M)-C did not meet the overall specifications required for use in the blast furnace, it was identified as a suitable reductant for other metallurgical applications.Item Characterisation and surface finish evaluation of Direct Energy Deposited AlCoCrCuFeNi High Entropy Alloys(University of the Witwatersrand, Johannesburg, 2024-01) Modikwe, Thembisile Patience; Mathe-Maleboho, Ntombi; Maledi, NthabisengThis study focused on the use of direct energy deposited (DED) techniques for the fabrication of AlCrCoCuFeNi high entropy alloy (HEAs) samples. HEAs have become a ground-breaking research field that provides solutions to complex problems in the aerospace industry. The industry requires improvement in the application of structural materials that are well-functioning at a low cost for example turbine blades. The fabrication of HEAs via DED commonly produces poor surface finish Ra in the range of 5 μm - 20 μm due to the layer-by-layer deposition method, as a result, it fails the industrial application requirements where the usual range of roughness tolerance required in the industry ranges from Ra is 0.8 μm < Ra < 1.6 μm thus, the need to deploy post-processing methods. This study focused on electropolishing (EP) and centrifugal barrel finishing (CBF) of AlCrCoCuFeNi-HEA samples. The polishing was performed using 80% methanol and 20% per-chloric acid solution used as the electrolyte. The samples were polished for 30 and 60 seconds in a Struers LectroPol-5 electrolytic polishing and etching device. The surface removal at 1200W for 30 sec on sample a was 50.29%, 58.65% for sample b, and 75.48% for sample c. The surface removal at 1400W for 60 sec on sample d is 63.25%, 45% for e, and 49.19% for f. The samples were polished for 7 and 14 hours in a CB320-CBF. During the period of 14 hours, a surface removal where the proportion of material removed for sample a was 55.37%, sample b was 43.13%, and sample c was 32.2% at a laser power of 1200W. After 7 hours of polishing, sample d achieved a surface removal of 86.02%, sample e achieved a surface removal of 43.18%, and sample f achieved a surface removal of 90% at a power of 1400W. Oxidation tests were conducted in static air at 1000˚C for 200h. The presence of FCC, BCC, and Fe2O3 oxide scales resulted in a noticeable increase in mass, with Fe2O3 scales being the most prevalent.Item Selection of a technique to separate carbon dioxide from methane for recovery of natural gas at Lake Kivu(University of the Witwatersrand, Johannesburg, 2024-02) Ntini, Hermann Ekini; Nkazi, Diakanua; Mukaya, ElieLake Kivu is situated between the Democratic Republic of Congo (DRC) and Rwanda. It is known to contain large amount of dissolved carbon dioxide and methane. It is termed a killer lake due to the toxic nature of these gases, which could emerge on the surface during a catastrophic eruption and cause massive devastation in this region. Extracting these toxic gases proves to be crucial to avoid natural disasters and to afford economic benefits in the form of electricity generation or energy export.Item Wear Reduction and Media Density Optimization for the Single Stage Pipe Densifier at Sishen Iron Ore’s Beneficiation Plants(University of the Witwatersrand, Johannesburg, 0202-02) Botha, Simone; Kabezya, KitungwaThe depleting high-grade iron ore mining supply at Sishen Mine in the Northern Cape, South Africa, has given rise to its beneficiation plants operating at higher media densities to upgrade lower-grade ore. In this study, densification was numerically modelled using an MPPIC model and experimentally tested using a 200-mm diameter centrifugal densifier from two local suppliers – Multotec and HMA. Shear stress, wear rate, separation efficiency and media losses were measured at increasing operating densities and differing vortex finder sizes. Optimum operating conditions were established. It was found that a feed density of 3.60 t/m3 and a shear stress of 9.70 e-3 N/m3 at the inlet using a vortex finder diameter size of 30 mm exhibited favourable performance in terms of media densification and downstream recovery. The practical significance of this is proven in terms of wear rate and its predictability to provide a consistent overflow of below 1.20 t/m3 media to the recovery circuit. Furthermore, information about ideal operating conditions in terms of inlet pressure and controls to identify premature failures were established.Item Investigating the effect of size, density and shape of Iron ore particles on batch jig performance(University of the Witwatersrand, Johannesburg, 2024-02) Dzaringa Kisembo, Daniel Elvis; Woollacott, LorenzoThe gravity separation method is one of the oldest methods of mineral beneficiation that takes advantage of the difference in the specific gravity of particles that are being separated. The separation occurs in a fluid medium, usually water, and involves floating off lighter material to leave behind denser ones. There are several types of gravity separation techniques, and they vary according to the equipment that is used for the separation or the property of the medium that is being used; the main gravity separation methods that are widely used for the beneficiation of Iron ore are Jigging and Dense Medium Separation (DMS). In this research, the jigging method is selected to investigate the concentration of an Iron ore by using a batch laboratory Jig; the jigging method was preferred for its simplicity and availability, generally Jigging has several advantages, some of which include cost effectiveness and simplicity of operation and its minimum impact on the environment. During the beneficiation of minerals using the jigging method of ore concentration, several feed material characteristics affect the efficiency, such as the particle density, size and shape. The aim of this research was to investigate the effect of these feed properties on jig performance. Tests were conducted on a Hematite ore sample using a batch jig to gain a deeper understanding of how the density, the size and shape of particles affect segregation. The iron ore samples were screened and any extremely small particles were removed, maintaining a particle size range between 2.8 and 10 mm. The results showed that particles were stratified on the basis of their specific gravity, denser particles reported toward the bottom layer of the bed and separated more efficiently. Less denser particles reported more toward the upper layer of the bed and were less efficiently separated. Coarser particles tend to report to the bottom layer of the bed and finer particles to the top product layer. Particles that were flatter and more elongated tended to end up in the bottom layer of the bed more often, while more rounded particles were not as likely to be found in the bottom layer.