Browsing by Author "Lokhat, David."

Now showing 1 - 20 of 33

A framework for modelling the interactions between biochemical reactions and inorganic ionic reactions in aqueous systems.
(2022) Brouckaert, Christopher John.; Lokhat, David.
Bio‐processes interact with the aqueous environment in which they take place. Integrated bio‐process and three‐phase (aqueous–gas–solid) multiple strong and weak acid/base system models are being developed for a range of wastewater treatment applications, including anaerobic digestion, biological sulphate reduction, autotrophic denitrification, biological desulphurization and plant‐wide wastewater treatment systems. In order to model, measure and control such integrated systems, a thorough understanding of the interaction between the bio‐processes and aqueous‐phase multiple strong and weak acid/bases is required. This thesis is based on a series of five papers that were published in Water SA during 2021 and 2022. Chapter 2 (Part 1 of the series) sets out a conceptual framework and a methodology for deriving bioprocess stoichiometric equations. It also introduces the relationship between alkalinity changes in bioprocesses and the underlying reaction stoichiometry, which is a key theme of the series. Chapter 3 (part 2 of the series) presents the stoichiometric equations of the major biological processes and shows how their structure can be analysed to provide insight into how bioprocesses interact with the aqueous environment. Such insight is essential for confident, effective and reliable use of model development protocols and algorithms. Where aqueous ionic chemistry is combined with biological chemistry in a bioprocess model, it is advantageous to deal with the very fast ionic reactions in an equilibrium sub‐model. Chapter 4 (part 5 of the series) presents details of how of such an equilibrium speciation sub‐model can be implemented, based on well‐known open‐source aqueous chemistry models. Specific characteristics of the speciation calculations which can be exploited to reduce the computational burden are highlighted. The approach is illustrated using the ionic equilibrium sub‐model of a plant‐wide wastewater treatment model as an example. Provided that the correct measurements are made that can quantify the material content of the bioprocess products (outputs), the material content of the bioprocess reactants (inputs) can be determined from the bioprocess products via stoichiometry. The links between the modelling and measurement frameworks, which use summary measures such as chemical oxygen demand (COD) and alkalinity, are described in parts 3 and 4 of the series, which are included as appendices to the thesis. An additional paper, presenting case study on modelling an auto‐thermal aerobic bio‐reactor, is included as a third appendix, as it demonstrates the application of some of the principles developed in the series of papers.
A generalized novel group contribution kinetic modeling approach to linear alkene metathesis.
(2023) Bansi, Nikhiel Isaiah.; Lokhat, David.
Traditional modeling approaches for linear metathesis systems involve the use of complicated approaches – This paper proposes a generalized novel kinetic modeling technique, in which the substituent (alkyl) groups on the transition states are postulated to have a unique effect on the rate at which the controlling step (transition state dissociation) occurs. Metathesis involves the redistribution of the carbon-carbon double bond across constituents, to create longer chain hydrocarbons, which significantly enhances product value in an industrial context. This work proceeds by unpacking the mechanics of the 1-Hexene (linear) metathesis pathway and utilizes this as a basis for developing a group contribution kinetic modeling approach. The substituent (alkyl) groups present in the forming transition state complexes were used to define (identify) rate constants that would potentially control the rate of formation of the various transition state complexes. A key assumption in this work was that the largest externally attaching substituent, and the substituent existing in the metallacyclobutane complex, be selected as the groups that define the formation of the respective transition state complexes. These observations resulted in a system of 25 uniquely defined (identified) rate parameters that were derived based on the defining groups observed within the mechanisms and are sufficient to describe the metathesis system under investigation. Literature experimental data, at a range of temperatures (420℃ – 460℃) was readily available and was used for the purpose of fitting the identified rate parameters, by simultaneously solving the system of differential equations that result from this system. Given the system size (25 parameters), and complexity of the interactions in the system, evolutionary and swarm optimization techniques were found to be fit for this purpose. It was found that combination between a genetic algorithm (GA) and particle swarm optimization (PSO) approach yielded identified parameters that minimized the overall error of the prediction. Kinetic parameters were identified, and Arrhenius plots were developed – These allowed for the activation energies (𝐸𝑎) and pre-exponential factors (𝐴𝑜) to be determined for each parameter. This was tested using literature experimental datasets, and predictions were found to present with a modified fitness value between 1.968 and 50.55. This illustrates that the novel group contribution kinetic modeling approach was suitable to define the interactions in the system. Further research is required to generalize this result over other alkene metathesis systems; however, this work proves that the approach is viable. Extending this work to ring closing metathesis (RCM), is a future area of research that is of particular importance, as pharmaceutical intermediates result from these processes, and represents the unique opportunity to create a synergistic industrial landscape.
Application of artificial neural networks as a predictive tool for the analysis of chemical engineering processes.
(2017) Khumalo, Siphesihle Praise-God.; Lokhat, David.
Abstract available in PDF file.
Bioethanol production from excess food crops in Nigeria: process design, optimization, and techno-economic analysis.
(2021) Awoyale, Adeolu Abiodun.; Lokhat, David.
The global drive for diversification of energy sources, particularly by focusing less on non-renewable fossil fuels and harnessing renewable energy resources like bioethanol, has motivated this research work. Before the last 10 years, bioethanol meant for use as fuel was produced from carbohydrate-rich crops such as cassava, yam, maize, millet, rice amongst others. Because the production of bioethanol from these food crops has been envisaged to jeopardize food security, the focus has been shifted to the production of bioethanol from the residues left behind after processing the food crops. These residues can be classified as lignocellulosic biomass. The major concentration of this study is the production of bioethanol from residues of food crops, namely, corn cobs, rice husks, sugarcane bagasse, cassava peels, and yam peels. The biomass used in this research were sourced from different locations in Nigeria, where they are found in abundance at certain seasons yearly. In the course of the work, the biomass were sieved into two mesh sizes of 300 and 425 microns, and also some of the biomass as well as all the five biomass were all mixed and firstly characterized to evaluate the effects of particle size as well as hybridized biomass mixtures on the end products and production efficiency of bioethanol. The effects of the adopted pretreatments in this study on the biomass were also investigated, as such, three types of pretreatments were adopted in this study namely; combined hydrothermal and acid pretreatment, combined hydrothermal and alkaline pre-treatment, and hydrothermal only pretreatment. The results of the characterization of the different biomass, including the hybridized biomass after pretreatment showed the pore features for hybridized corn cobs and rice husks biomass have the maximum specific surface area and pore volume of 1837 m2/g and 0.5570 cc/g respectively. Also, the values of the cellulose content improved slightly with the pretreatment and the value of the lignin content decreased considerably. The cellulose values range from 34.2 to 36.5 wt% for the acid, alkali and hot water pretreated hybridized biomass. Releases from the pretreatment process to air, soil, and water were measured with SimaPro. The environmental impact categories accessed include global warming potential (GWP)/climate change, and acidification (AP). With a mean value of 15.82 kg CO2 (eq), the alkaline pretreatment using sodium hydroxide shows the highest release of GHG emissions, while acid pretreatment employing dilute sulphuric acid generated a mean value of 8.68 kg CO2. Hybridized feedstocks of cassava peels plus yam peels, and corn cobs plus rice husks biomass, were optimized using the Response Surface Methodology (RSM) centred on the statistical design of experiments (DOE) of the Box-Behnken design (BBD), in the production of bioethanol. The BBD was harnessed using a 3-level, 3-factor process variables using pH, time, and particle size. The bioethanol yield from the two hybridized biomass feedstocks was predicted by the developed quadratic polynomial models from BBD. The hybridized rice husks plus corn cobs biomass with a maximum bioethanol yield of 160 ml/1500 g biomass gave a better prospect for bioethanol production when compared with hybridized cassava peels plus yam peels biomass with a maximum bioethanol yield of 125 ml/1500 g biomass. This reinforces the finding that hybridizing the feedstocks enhances the capacity for better bioethanol yield after fermentation. The economic analysis of the produced bioethanol gave a price of 0.41 USD/l, which is a good deal as it compares favorably well with the 0.45 USD/l price of ethanol in the Nigerian open market.
Catalytic hydrocracking of waste vegetable oil transition metal-based catalysts: selective production of jet fuel range alkanes.
(2015) Moodley, Keldon.; Lokhat, David.
The performance of transition metal-based catalysts on amorphous supports has been investigated for the high pressure (120 bar) hydrocracking of waste vegetable (cooking) oil in a fixed-bed tubular reactor between 400-450 °C. The study focused on the effect of the operating parameters (reaction temperature, type of transition metal catalyst and amorphous support, the sulphided state of the catalyst and the use of regenerated catalyst) on the yield of transportation fuel n-alkanes (C5-C18), and the primary product, kerosene (jet fuel) range n-alkanes using the One-Variable-At-A-Time approach. The objectives included characterising the feedstock and catalysts, determining the optimum catalyst and operating conditions to produce kerosene range n-alkanes and to estimate activation energies through a reaction kinetic study. Comparative studies of the results from commercially produced Ru/Al2O3 and Ni/Al2O3 catalysts and laboratory prepared Ni-Mo/SiO2 and Co-Mo/SiO2 were undertaken. All tested catalysts were effective in achieving kerosene range n-alkanes in the liquid product while achieving oil conversions > 62 wt.%. While the laboratory catalysts only had liquid and gas products, the commercial catalysts also had a waxy residue product indicating a lower hydrotreating activity on the metallic sites. Furthermore, both laboratory catalysts had higher n-alkane yields indicating a higher activity for hydrocracking reactions on the acidic sites on the SiO2 than the acidic sites on the Al2O3 support. The best yield of kerosene range n-alkanes obtained from the experimental design was 5.84 wt.% using the fresh Ni-Mo/SiO2 catalyst at 450 °C, with an oil conversion of 90.32 wt.%. An increase in oil conversion and liquid n-alkane products with an increase in reaction temperature for all tested catalysts indicated that hydroprocessing (hydrotreating and hydrocracking) reactions are favoured at higher temperatures. Furthermore, sulphiding the catalyst prior to use was found to greatly increase the catalyst activity in promoting hydroprocessing reactions. Results from the use of regenerated catalysts show a small decrease in the yield of kerosene range n-alkanes when compared to the corresponding fresh catalyst. This suggests that regeneration of the spent catalyst and subsequent re-use may be a feasible option. A simple kinetic model of the hydroprocessing reactions was developed and kinetic parameters were identified by regression of the experimental data. The regenerated Ni/Al2O3 catalyst had the lowest estimated activation energy of 14.37 kJ/mol, while the fresh Ni-Mo/SiO2 catalyst had the highest estimated activation energy of 51.75 kJ/mol from the studied catalysts.
Characterization of fluidization regimes by analysis of pressure fluctuations in gas-solid fluidized beds.
(2017) Naidoo, Sayuri.; Lokhat, David.
Fluidized beds are ranked as the top contacting method with the best overall benefits, and have been used over many years in several industrial applications. Literature indicates that pressure fluctuations are influenced by variables related to fluidization regimes in a fluidized bed; such as the bubble size, bubbling rising velocity and the motion of the bed surface (Fan et al., 1981). Hence several researchers have employed pressure fluctuations to aid in the understanding of fluidized bed system hydrodynamics. This study was focused on gas-solid fluidized beds, during aggregative fluidization represented by the bubbling, slugging and turbulent regimes. Geldart (1973) materials from the classification were studied in this research; Group A (spent Fluid Cracking Catalyst), Group B (sand) and Group D (plastic beads). The experimental equipment was composed of an existing laboratory-scale gas-solid fluidized bed and data acquisition system. Three transparent fluidized bed columns were investigated; fluidized bed 1 (I.D 5 cm), fluidized bed 2 (I.D 11 cm) and fluidized bed 3 (I.D 29 cm). The time-series analysis of pressure fluctuation signals were investigated using the time and frequency domain methods. The pressure fluctuation signal was converted into the frequency domain by use of the Fast Fourier Transform (FFT). For increased bed heights the power spectrum was narrower, higher in amplitude, had more distinct peaks and the dominant frequency was lower, when compared to the lower bed height for the same material and fluidization regime. Also decreasing dominant frequencies and large increases in the amplitude of the pressure fluctuation were observed for each increasing fluidization regime; from the bubbling to slugging and to the turbulent regimes. The research contribution from this study was realized, as a range of dominant frequencies were successfully identified for each specific fluidization regime at its respective velocities. The identification of the transition phase was accomplished with low accuracy from the research contribution. It was recommended to employ differential pressure measurements for larger columns to increase the accuracy of data achieved; thereby permitting the comparison of useable power spectra results for scale-up.
Development of a continuous process for the production of hexafluoropropyl methyl ether.
(2015) Domah, Ashveer Krishen.; Lokhat, David.; Ramjugernath, Deresh.
Partially fluorinated dialkyl ethers are valuable intermediates for organofluorine syntheses. These compounds can be used for the preparation of perfluoroacrylic acids or the anhydrides, amides and esters thereof. They also serve as very effective solvents, particularly for the extraction of essential oils. A continuous process for producing 1,1 ,2,3,3,3-hexafluoropropyl methyl ether by reacting a liquid mixture of potassium hydroxide and methanol with gaseous hexafluoropropene in one or more microstructured devices was developed. The reaction of hexafluoropropene and potassium methoxide is highly exothermic, with higher operating temperatures favouring the formation of hexafluoropropyl methyl ether. The reactants are contacted for a prescribed time within a reaction zone having a high heat transfer area to reaction volume ratio and in intimate contact with a cooling medium, facilitating efficient dissipation of the exothermic reaction heat. The product mixture is contacted with water at below ambient temperature to extract the residual methanol and the raffinate is further purified by means of conventional distillation. The water and methanol mixture is fed to a distillation column that recovers methanol at a purity of 99%. The HME-methanol mixture is fed to another distillation column which produces HME at 98% purity. The methanol recovered from both distillation columns is recycled to the start of the process. The synthesis of hexafluoropropyl methyl ether using the aforementioned process was demonstrated experimentally using a falling film microreactor. Quadratic response surface methodology was used to probe for optimal reaction conditions for the yield of hexafluoropropyl methyl ether as well as the purity of the raw product.
Development of the sugarcane biorefinery economic analysis toolbox (S-BEAT) as a product and process selection support for South African sugar mills.
(2020-12) Naidoo, Prelene.; Stark, Annegret.; Lokhat, David.
In recent years, key stakeholders within the South African Sugar Industry have realised the necessity to diversify their product portfolios further and move from only sugar-based products to the production of chemicals, materials, and fuels. Numerous value-added products can potentially be generated from the available streams in a sugar mill. In theory, these products can supply numerous markets of various sizes and hence generate new revenue streams for the sugar mill. A toolbox, the Sugarcane Biorefinery Economic Analysis Toolbox (S-BEAT) which uses a preliminary cost estimation method, was developed to serve this purpose. S-BEAT provides both a cost estimation and economic analysis at a preliminary process design stage for preselected product and process alternatives in the South African context. S-BEAT makes use of the order of magnitude approach, which is based on data from existing plants. It accepts historical data and escalates the capital investment to the current year, whilst making adjustments for differing product capacities and plant locations. It is estimated that this method has an accuracy of about ±30% to ±50%, which is considered satisfactory for a preliminary cost estimate. The cost estimates then undergo an economic analysis to determine product profitability (Net Present Value (NPV), Internal Rate of Return (IRR) and Discounted Payback Period). S-BEAT allows for a comparison between process and product alternatives, as well as between raw material type (clear juice, mixed juice, syrup, A-sugar, or A-molasses) and quantity (effects of economies of scale) of a sugar mill stream diverted from sugar production to the biorefinery operation. Furthermore, it comprises a sensitivity analysis option which is used to identify those process variables that exhibit the largest effect on the overall economics, serving as pointers for specific process optimisation. With this toolbox, a techno-economic analysis of several products, including high-density polyethylene (HDPE), polylactic acid (PLA), monoethylene glycol, lysine, succinic acid, 1,4- butanediol and 2,5-furandicarboxyllic acid, was conducted. The economics, in the form of the NPV, IRR and Discounted Payback Period, from the aforementioned raw material streams were assessed. Additionally, the Minimum Selling Price that is required to ensure a plant is feasible is included for a range of feedstock costs and numerous plant capacities. The results of this assessment are presented comparatively and provide a basis for decision making for the local sugar industry on which processes and products should be further investigated at a more rigorous techno-economic modelling level. The S-BEAT toolbox is available to the South African Sugar Industry to assess products using in-house specific data. Further products and feedstock streams can be added, to support decision making at this early stage of the industry’s transformation to a diversified industry.
Exploring the effectiveness of separation of pith/fibre fractions in sugarcane bagasse briquetting.
(2019) Madlala, Nkosinathi Emmanuel.; Eloka-Eboka, Andrew Chukwudum.; Lokhat, David.
Abstract available in the PDF.
Gas-phase epoxidation of hexafluoropropene over supported copper oxide catalysts: mechanism and kinetics.
(2021) Ndlovu, Lindelani Archie.; Lokhat, David.; Ramjugernath, Deresh.
Abstract available in PDF.
Gas-phase epoxidation of hexafluoropropylene.
(2012) Lokhat, David.; Starzak, Maciej.; Ramjugernath, Deresh.
A process was developed for the production of hexafluoropropylene oxide via the gas-phase oxidation of hexafluoropropylene with molecular oxygen. The non-catalytic oxidation reaction was investigated in an isothermal, laminar flow reactor at a total pressure of 4.5 bar and over the temperature range of 453 to 503 K. Feed mixtures comprising 20 to 67% HFP in oxygen were used at total flow-rates between 150 and 550 cm3 min-1. The reactor was fabricated from 1/8 inch nominal sized, copper refrigeration tubing and was 114.3 m long. The reactor was used in the form of a helical coil. Gas-chromatographic analysis was used for reactant and stable product quantification. The reaction gave hexafluoropropylene oxide, trifluoroacetyl fluoride and carbonyl fluoride as major products. Minor products included tetrafluoroethylene and hexafluorocyclopropane. The oxidation reaction also produced high molecular weight oligomers that were retained on the inner surface of the reactor tube. The operating conditions for the non-catalytic oxidation were optimized independently for HFPO selectivity and yield using quadratic response surface methodology. A maximum HFPO selectivity of 55.81% was identified at 478.2 K, a HFP/O2 molar feed ratio of 1.34 mol mol-1 and a space time of 113 seconds. An optimum HFPO yield of 40.10% was identified at 483.2 K, a HFP/O2 molar feed ratio of 1.16 mol mol-1 and a space time of 121 seconds. Using the weighted-sum-of-squared-objective-functions (WSSOF) multi-response optimization technique, a combined optimum HFPO selectivity and yield of 56% and 40%, respectively, was obtained at 480 K, with a HFP/O2 molar feed ratio of 1.21 and a space time of 118 seconds. This represented the best trade-off between these two performance criteria. A kinetic reaction scheme involving 8 species and 7 reactions was developed, based on the results of the experimental study, and was used to model the non-catalytic oxidation of HFP. The initial steps in this scheme encompassed the addition of oxygen to the double bond of the fluoro-olefin and transformation of the resultant dioxetane intermediate to form HFPO and the haloacetyl fluorides. Subsequent steps included the thermal decomposition of HFPO to yield CF3COF, C2F4 and c-C3F6, as well as elimination of C2F4, and to a lesser extent CF3COF, through oxidation. Rate parameters for the oxidation reactions were determined through a least-squares minimization procedure. The investigation was extended by considering the catalyzed synthesis of HFPO. Four different catalysts were studied viz., 1wt% Au/A12O3, 1wt% Au/ZnO, 10wt% CuO/SiO2 as well as 10wt% CuO/SiO2 doped with caesium. The gold-based catalysts were found to be completely inactive for the oxidation reaction. The caesium promoted, copper-based catalyst appeared to be the most stable and active, with no observable decomposition to copper fluoride. At 453 K, a HFP/O2 molar feed ratio of 0.86 and a weight-hourly-space-velocity of 0.337 h-1, a HFPO selectivity of 85.88% was obtained. This was significantly greater than what was achieved in the non-catalytic system.
High temparature flue gas desulfurization: experiments and modelling.
(2017) Moodley, Calvin.; Lokhat, David.
As stringent environmental regulations regarding SO2 emissions have been enacted in many countries, the removal of SO2 from flue gas has become a necessity. Conventional methods for flue gas desulfurization (SO2 removal) involve cooling of the flue gas followed by scrubbing and subsequent reheating of the treated gas. To minimize cooling/heating requirements, thus optimizing energy consumption and related expenses, high temperature flue gas desulfurization technologies utilizing dry sorbent injection have recently been emphasized. Experimental studies reported in literature dictate that possible formation of sulfur trioxide (SO3) in situ has been largely ignored for the high temperature desulfurization process. Since the performance of desulfurization systems is based on the amount of SO2 removed, any conversion to SO3 would result in an overestimated efficiency with respect to SO2 removal. This study focused on determining if significant amounts of SO3 are formed at the excessive temperatures of high temperature flue gas desulfurization and what effect this has on the actual amount of desulfurization achieved. The experiment was designed with one independent variable (desulfurization temperature) and three levels for that variable. The aim of the experimental design was to evaluate the effect of desulfurization temperature (700oC, 800oC and 900oC) on the amount of SO2 removed using limestone as the solid sorbent – through determination of breakthrough time and sorbent sorption capacity (based on saturation time) – and the amount of SO3 formed – estimated via the Isopropanol Absorption Bottle Method. Reaction of SO2 within a packed bed of limestone sorbent particles was modelled by considering two descriptions with regards to the morphological structure of the sorbent surface: a non-porous surface and a porous surface. Solution of the developed model was undertaken utilizing the Method of Lines (MOL). Experiments were carried out using a bed height of 3 cm and fixing the inlet flowrate of flue gas at 6Lmin, resulting in a residence time within the sorbent bed of 0.3 seconds. A gas chromatograph (GC) was utilized to generate the transient SO2 concentration profile at the reactor inlet and outlet. This ultimately determines breakthrough time, saturation time and sorbent sorption capacity. The GC was thus required to be calibrated. A third-degree polynomial was found to best fit the calibration curve data. A dead time of three seconds was estimated for flue gas to propagate through the system to the sample extraction point. Sieve tray analysis of the limestone sorbent particles revealed an average particle size of 362.76μm. Experimental results for SO2 removal indicate that breakthrough time, saturation time and SO2 sorption capacity of limestone sorbent all increase with an increase in operating temperature. This trend is attributed v to formation of larger volume product CaSO4 layers on active sorbent surfaces effecting a more gradual decline in overall diffusivity (Dz) with increasing operating temperature, as expressed by results of the model analysis. Breakthrough times of 100s, 210s and 240s were achieved for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively; saturation times of 890s, 1060s & 1200s were achieved for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively; SO2 sorption capacities of 13.50mg ofSO2gram of Sorbent, 16.20mg ofSO2gram of Sorbent and 18.73mg ofSO2gram of Sorbent were achieved for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively. The model analysis proves further that the limestone sorbent particles are porous in nature due to a conservative fit between experimental data and the predicted model solution. Results for SO3 formation indicate that the generation of SO3 is predominantly due to the heterogeneous catalytic reaction of SO2 with the stainless steel walls of the reactor at such elevated temperatures utilized during high temperature flue gas desulfurization. In the absence of stainless steel the mass of sulfur entering the system which is converted to SO3 was calculated to be 0.251%, 0.249% and 0.247% for the operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively. In the presence of stainless steel the mass of sulfur entering the system which is converted to SO3 was calculated to be 3.24%, 5.60% and 9.30% for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively. It can thus be concluded that as operating temperature shifts away from 700oC towards 900oC the formation of SO3 becomes much more significant in the presence of stainless steel. The amount of desulfurization achieved however is similar for the respective operating temperatures. Since stainless steel is a requirement it is recommended to operate at a temperature of 700oC since similar performance is achieved by the limestone sorbent in terms of SO2 removal, relative to 800oC and 900oC, while SO3 production is relatively minimal.
High temperature flue gas desulfurization: experiments and modelling.
(2018) Moodley, Calvin.; Lokhat, David.
As stringent environmental regulations regarding SO2 emissions have been enacted in many countries, the removal of SO2 from flue gas has become a necessity. Conventional methods for flue gas desulfurization (SO2 removal) involve cooling of the flue gas followed by scrubbing and subsequent reheating of the treated gas. To minimize cooling/heating requirements, thus optimizing energy consumption and related expenses, high temperature flue gas desulfurization technologies utilizing dry sorbent injection have recently been emphasized. Experimental studies reported in literature dictate that possible formation of sulfur trioxide (SO3) in situ has been largely ignored for the high temperature desulfurization process. Since the performance of desulfurization systems is based on the amount of SO2 removed, any conversion to SO3 would result in an overestimated efficiency with respect to SO2 removal. This study focused on determining if significant amounts of SO3 are formed at the excessive temperatures of high temperature flue gas desulfurization and what effect this has on the actual amount of desulfurization achieved. The experiment was designed with one independent variable (desulfurization temperature) and three levels for that variable. The aim of the experimental design was to evaluate the effect of desulfurization temperature (700oC, 800oC and 900oC) on the amount of SO2 removed using limestone as the solid sorbent – through determination of breakthrough time and sorbent sorption capacity (based on saturation time) – and the amount of SO3 formed – estimated via the Isopropanol Absorption Bottle Method. Reaction of SO2 within a packed bed of limestone sorbent particles was modelled by considering two descriptions with regards to the morphological structure of the sorbent surface: a non-porous surface and a porous surface. Solution of the developed model was undertaken utilizing the Method of Lines (MOL). Experiments were carried out using a bed height of 3 cm and fixing the inlet flowrate of flue gas at 6Lmin, resulting in a residence time within the sorbent bed of 0.3 seconds. A gas chromatograph (GC) was utilized to generate the transient SO2 concentration profile at the reactor inlet and outlet. This ultimately determines breakthrough time, saturation time and sorbent sorption capacity. The GC was thus required to be calibrated. A third-degree polynomial was found to best fit the calibration curve data. A dead time of three seconds was estimated for flue gas to propagate through the system to the sample extraction point. Sieve tray analysis of the limestone sorbent particles revealed an average particle size of 362.76μm. Experimental results for SO2 removal indicate that breakthrough time, saturation time and SO2 sorption capacity of limestone sorbent all increase with an increase in operating temperature. This trend is attributed v to formation of larger volume product CaSO4 layers on active sorbent surfaces effecting a more gradual decline in overall diffusivity (Dz) with increasing operating temperature, as expressed by results of the model analysis. Breakthrough times of 100s, 210s and 240s were achieved for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively; saturation times of 890s, 1060s & 1200s were achieved for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively; SO2 sorption capacities of 13.50mg ofSO2gram of Sorbent, 16.20mg ofSO2gram of Sorbent and 18.73mg ofSO2gram of Sorbent were achieved for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively. The model analysis proves further that the limestone sorbent particles are porous in nature due to a conservative fit between experimental data and the predicted model solution. Results for SO3 formation indicate that the generation of SO3 is predominantly due to the heterogeneous catalytic reaction of SO2 with the stainless steel walls of the reactor at such elevated temperatures utilized during high temperature flue gas desulfurization. In the absence of stainless steel the mass of sulfur entering the system which is converted to SO3 was calculated to be 0.251%, 0.249% and 0.247% for the operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively. In the presence of stainless steel the mass of sulfur entering the system which is converted to SO3 was calculated to be 3.24%, 5.60% and 9.30% for operating temperatures of 700 ℃, 800 ℃ and 900℃ respectively. It can thus be concluded that as operating temperature shifts away from 700oC towards 900oC the formation of SO3 becomes much more significant in the presence of stainless steel. The amount of desulfurization achieved however is similar for the respective operating temperatures. Since stainless steel is a requirement it is recommended to operate at a temperature of 700oC since similar performance is achieved by the limestone sorbent in terms of SO2 removal, relative to 800oC and 900oC, while SO3 production is relatively minimal.
Highly dispersed zinc-based sorbents for desulphurisation : synthesis and application.
(2016) Govender, Neelan.; Lokhat, David.; Carsky, Milan.
Abstract available in PDF file.
Hydrocracking of short residue over unsupported and supported magnetite nanocatalysts.
(2017) Maharaj, Shaheel.; Lokhat, David.
With the rapid depletion of crude oil and current cracking methods of heavy petroleum residue all resulting in the production of undesirable coke formation, a better solution must be found. This project investigated the use of an unsupported molybdenum-doped magnetite nano-catalyst, as well as a magnetite nanocatalyst on a mesoporous silica support, to determine if the use of these catalysts can be successful in cracking petroleum residue. Short residue from the vacuum distillation column supplied by SAPREF, was used throughout the experimental work. A lot of effort went into the preparation of the feedstock due to the high viscosity of short residue. The solvent used during experimental work was toluene, which was used to dilute the short residue. A temperature range between 350˚C and 400˚C was used in order to determine temperature effects on product distribution from the cracking reaction. The feedstock to catalyst ratio was also varied, using the unsupported catalyst, in order to determine the effects of the amount of catalyst on the reaction. Kerosene and gas oil are the desired products due to their higher heating value and use as liquid fuels compared to the heavier residue. There is a strong interaction between temperature and catalyst to feedstock ratio. The high temperature-high catalyst combination gave improved gas oil yields over the low temperature-high catalyst combination. Results carried out at 400˚C with a high catalyst amount showed the most favourable results with a yield of 49.3% and 6% of gas oil and kerosene respectively. Aquaprocessing (catalytic splitting of water that occurs on the surface complexes of the iron-based catalyst, at a relatively low pressure) was simulated at the experimental conditions using kinetics from literature for a nickel-based catalyst. The simulated composition profiles proved that the unsupported magnetite nanocatalyst was much more efficient in upgrading residue than the nickel based catalyst, due to the presence of greater amounts of lighter components. Analysis of the catalyst after the cracking reaction shows that no major phase changes had taken place and that the catalyst could be regenerated to be used again. The supported magnetite nanocatlyst was compared to conventional nickel-molybdenum and cobalt-molybdenum catalyst, in a fixed bed reactor set up. The supported catalyst proved to be the most consistent, and was able to shift the residue into the lighter fractions more effectively than the conventional catalysts. The supported catalyst was the most effective in cracking the vacuum residue, mostly into vacuum gas oil. The yields using the catalyst compared quite favourably with the unsupported catalyst, with the unsupported catalyst yielding more lighter components. The most favourable results implementing a supported catalyst were also at 400˚C, due to the extensive decrease in vacuum residue and a corresponding increase in lighter components. Ultimately this investigation proved that hydrocracking can take place with the use of a supported and unsupported magnetite nanocatalyst, at lower temperatures than that of conventional methods and aquaprocessing. It was also proven that the process can be upscaled to industry level, as shown with the performance of the supported catalyst. A larger temperature range could give better clarity in the performance of the catalyst for future petroleum residue cracking.
Hydrothermal liquefaction of marine macroalgae.
(2021) Nadar, Deslin.; Lokhat, David.
The biofuel industry has experienced substantial growth during the past decade due to the extreme demands placed on the fossil fuel industry and the limited availability of fossil fuels. Biofuels are seen as a renewable source of energy while reducing the effects on the environment significantly. Renewable biofuels are made through the use or conversion of biomass such as algae and lignocellulosic biomass. Biomass is seen as a viable alternative to produce biofuel as it is readily available, and has a relatively low cost. Marine macroalgae (seaweed) may be considered as a feedstock for biofuel production due to their low cost, fast growth rate, and they do not cause land-use and fuel-vs-food conflicts. Hydrothermal liquefaction is a thermochemical process that utilises water as a reaction medium under high pressures and temperatures to produce bio-oils from biomass. Hydrothermal liquefaction is different from most other conversion techniques as it uses a wet feedstock and does not require an energy-consuming drying step. In this work, hydrothermal liquefaction of marine macroalgae for the production of bio-oil was studied at various reaction conditions. The effect of the mass of seaweed, temperature, pressure, solids loading and reaction time were examined. A kinetic model of dissolution was developed and regressed against the experimental temporal data to obtain the kinetics of dissolution. A measured quantity of marine macroalgae and water were placed within the Parr reaction vessel and exposed to high temperatures and pressures for a set time. The resulting solution was filtered, to separate the algae from the liquid (water and bio-oil solution), and mixed with dichloromethane, to selectively separate the bio-oil from the water. The dichloromethane mixture was transferred to the rotary evaporator and the dichloromethane was evaporated to ensure only the bio-oil remained. The bio-oil was measured and transferred to the GC/MS for a more in-depth compositional analysis. Bio-oil was formed for every variation of the process variables and every run conducted. The highest bio-oil yield obtained was for the 10g 10wt% run at the high reaction conditions (250°C and 4000 KPa) and a time of 30 minutes, with a bio-oil yield of 34.67%. This was for the highest manipulation of every process variable. The lowest bio-oil yield (not including the induction period) was obtained for the 6g 10wt% run at the low reaction conditions (200°C and 1500 KPa) and a time of 5 minutes, with a bio-oil yield of 18.14%. The bio-oil yield formed during the induction period ranged from 0.11% to 26.58%. A higher mass loading was observed to provide a higher dissolution and a higher bio-oil yield (ranging from 29.59% to 34.67% for a mass loading of 10g)). Higher temperatures and pressures were also found to increase the mass dissolution and bio-oil yield obtained. The higher solids loading of 10wt% observed a larger bio-oil yield (ranging from 27.96% to 32.62%) than a solids loading of 5wt% (ranging from 22.81% to 26.53%). The bio-oil yield was found to increase for an increase in the reaction time for every variation of the process variable. The assessment of the quality of bio-oil through GC/MS analysis determined that the main compounds formed during the hydrothermal liquefaction process were hexanedioic acid (adipic acid), cyclopentene, hexadecenoic acid, phenol, butanone, ethanone, tetrapentacontane, furancarboxaldehyde, cyclohexane, and hexanedioic acid- bis (2-ethyhexyl) ester. A kinetic model was applied to the data obtained to determine the kinetic parameters of dissolution. The dynamic model was identified with the aid of MATLAB programming software. The kinetic models for the conversion of solids to bio-oil and the conversion of solids to the aqueous product have the same formula. The simplified model is expressed by the mass fraction of the solid biomass multiplied by the kinetic rate constant and then multiplied again by the exponential of the negation of the inhibition constant over the mass fraction of the solid biomass. Utilising both the non-linear least squares regression and the ode15s variable-step, variable-order solver, the kinetic reaction rates were determined to be 0.0059 g/g/s (𝑘1) for the conversion from solids biomass to bio-oil and 0.0103 g/g/s (𝑘2) for the conversion from solid biomass to the aqueous-phase product. The inhibition constants (𝑘3 and 𝑘4) were determined to be the same at a value of 4.44e-14. The overall results of this work validate that the hydrothermal liquefaction of marine algae produces an adequate amount of bio-oil that may be further processed to produce biofuel. It was observed that higher process conditions resulted in higher bio-oil yields being obtained and that a kinetic model may be determined for the mass dissolution from the algae and bio-oil yield formed. The maximum yield of 34.67% obtained in this work was amongst the higher yield results for research in this section, while utilizing lower temperatures and a slightly higher reaction time, thereby requiring a lower amount of energy. The results of this work imply that enough bio-oil is formed from the hydrothermal liquefaction of marine macroalgae to allow for scale-up of the process to produce a cleaner biofuel fuel that may alleviate the demands placed on fossil fuel
Investigation and optimization of intensified separation processes: treatment of aqueous organic mixtures using reactive extraction and emulsion liquid membrane techniques.
(2020) Inyang, Victoria Malachy.; Lokhat, David.
The application of intensified processes of reactive extraction and emulsion liquid membrane technique for the separation of various low molecular weight carboxylic acids (propionic acid, malic acid and butyric acid) from very dilute aqueous solution was undertaken. The aqueous phase feed concentration ranges of the different carboxylic acids for reactive extraction [propionic, butyric (0.4-1 kmol/m3), malic (0.1-1kmol/m3) acids] and emulsion liquid membrane (propionic and malic acid (0.05 – 0.1 k mol/m3) were chosen to simulate the actual aqueous waste streams and fermentation conditions encountered in industry. Trioctylamine extractant in 1-decanol as active diluent was used as the extractant (organic) phase to perform the experiment. The effect of different process variables on the extraction efficiency expressed in terms of distribution coefficient (KD) and degree of extraction (%E) was systematically determined. Three independent process variables were chosen, including temperature (298.15-313.15 K for propionic acid and malic acid and 298.15-318.15 K for butyric acid), initial organic acid concentration in the aqueous phase and trioctylamine composition (10-30 %) in the organic phase for the reactive extraction technique. The interactive effects and optimum values of these process design variables were determined using response surface methodology (RSM) for the reactive extraction process. The statistical design analysis demonstrated that the acid concentration and trioctylamine composition had significant effect while temperature had an insignificant effect on the response value as well as interactive and quadratic effect on the response. The optimum solution led to an experimentally determined extraction efficiency of 89.79% for propionic acid, 93.25% for malic acid and 96.45% for butyric acid. The extraction efficiency in the emulsion liquid membrane process is dependent on the studied parameters such as initial acid concentration, sodium carbonate concentration, trioctylamine concentration, treat ratio and extraction time. The formulation of the liquid membrane consists of trioctylamine as carrier, 1-decanol as modifier, sorbitan monooleate (Span 80) as surfactants in heptane and sodium carbonate (Na2CO3) as a stripping agent. Response surface methodology (RSM) and artificial neural network (ANN) was employed for experimental design, optimization, construction and interpretation of response/output surface plots so as to show the effect of input variables on extraction efficiency in addition to the combined effects between variables. The optimum solution achieved by RSM led to an experimentally determined extraction efficiency of 92.28% and 85.91% in the propionic and malic acid extraction respectively by (emulsion liquid membrane) ELM process. The intrinsic kinetic studies of reactive extraction were determined for propionic and malic acid extraction using dilute solutions of the acids with concentration range of 0.2 to 0.6 kmol/m3 and trioctylamine (10%v/v) in 1 decanol as extractant at 303.15 K. The kinetic process parameters such as reaction order, mass transfer coefficient and rate constant were evaluated using the experimental data. From the results obtained, the reaction was found to be an instantaneous second-order chemical reaction occurring in the organic diffusion film. The values of the rate constants were found to be 0.430 m3/mol s and 0.332 m3/mol s respectively for propionic acid and malic acid while the mass transfer coefficient, km was also obtained for propionic acid (9 x 10-6 m/s) and malic acid (3x10-6 m/s). From the results obtained, these intensified technique represents an effective method for the recovery of low concentrations of carboxylic acids from aqueous waste streams and fermentation broths, with emulsion liquid membrane offering significantproperties/characteristics like small quantity of organic phase and extractant, very fast extraction time, increased solute transfer rate and selectivity through the membrane, high selectivity and applicability in specie removal from very low to high concentrations and governed by a non-equilibrium mass transfer. It is therefore worth investing in this process or alternatively a hybrid of both reactive extraction and emulsion liquid membrane processes.
An investigation into the viscosity of C-massecuite using a pipeline viscometer.
(2017) Shah, Shaista.; Lokhat, David.; Peacock, Stephen David.
Sugar is recovered by three stages of evaporative crystallisation, with each stage producing a two-phase mixture of sugar crystal and mother liquor commonly referred to as massecuite. The composition of each massecuite changes as increased amounts of sucrose crystallises out of solution, increasing the concentration of residual non-sugars and organic salts with each evaporative stage and resulting in C-massecuite possessing the lowest purity and highest viscosity. Tongaat Hulett maintains an interest in the viscosity of C-massecuite from a process and equipment design perspective as viscosity is a critical physical property in the selection of pumps and design of piping networks, evaporative and cooling crystallisers, crystalliser drives and reheaters in the C-station of a sugar factory. In the absence of a well-established correlation, viscosity data published by Andre Rouillard in 1984 is widely-used at Tongaat Hulett, however, this chart and other widely published data resulted from experimentation with a rotating viscometer that is believed to be unsuitable for this application. The rotating viscometer results in displacement of sugar crystals, interfering with the accuracy of the measurements. Whilst the rotational viscometer was accepted by the International Commission for Uniform Methods of Sugar Analysis as the standard technique for measurement of molasses viscosity, no standard technique is available for massecuite viscosity measurement. An investigation into alternative methods of viscosity measurement rendered the pipeline viscometer as best suited to this product as the method of measurement is not affected by the heterogeneous nature of the massecuite. The aim of this study is thus to design, construct and validate a pipeline viscometer which is to be used, together with non-Newtonian theory, to investigate the viscosity of C-massecuite. The pipeline viscometer was successfully constructed, validated and used, together with the power law model, to describe the viscosity of C-massecuite in terms of two rheological parameters; the flow behaviour index and consistency. The results of this study indicate that the average flow behaviour index of C-massecuite is 0.85. An empirical correlation for C-massecuite consistency as a function of temperature, dissolved solids concentration and crystal content was proposed with a regression coefficient of 0.7672 as well as additional equations to guide the estimation of C-massecuite viscosity. The massecuite consistency, assumed to be equivalent to apparent viscosity at a shear rate of 1s-1, was compared with the C-massecuite viscosity data currently used. A more rapid increase in massecuite viscosity with a reduction in temperature was found, however, the experimental data was found to fall within the recommended range for C-massecuite viscosity currently used. It is with confidence that the power law model can thus be used with a flow behaviour index of 0.85 and a consistency as predicted by the empirical correlation and guiding equations to yield an apparent viscosity for C-massecuite.
An investigation of a two-step, temperature-staged, direct coal liquefaction process.
(2015) Singh, Reyna.; Lokhat, David.; Carsky, Milan.
From its inception in the 1700’s, deriving fuel from the Direct Coal Liquefaction (DCL) process has spawned numerous pursuits. While coal is an abundant fossil fuel in many countries and represents approximately 70% of the world’s total energy reserves (Birol, 2004), the DCL process is synonymous with the use of severe operating conditions and catalysts of poor activity. This work is an investigation of a two-step, temperature-staged DCL process and aimed at producing a high value liquid hydrocarbon product at, relatively, mild operating conditions. This stepwise process was initially carried out in a batch reactor. In this first stage, the aim was to maximise on the liquid product (oil) yield by enhancing the thermal dissolution of high grade bituminous type coal in tetralin as the hydrogen donor solvent, using 2:1 and 3:1 solvent: coal ratios. The oil obtained was refined by hydrotreating in a catalytic fixed bed reactor. Both stages were carried out isobarically at 100 barg and, in the first stage, temperatures of 250 ℃ and 300 °C were used. Thereafter, operating temperatures were staged with a 50 °C increase in the second stage reactor. In the first stage, molybdenum doped magnetite was used as the catalyst. The performances of cobalt-molybdenum (Co-Mo) and nickel-molybdenum (Ni-Mo) were trialled in the second-stage reactor. In order to assess the potential value of the oil between the stages, the oil was analysed using Gas Chromatography –Mass Spectrometry (GC-MS). Within the actual experimental boundary; oil yield, alkane and cycloalkane selectivity response data was fitted to linear models. In the first stage the liquid yield was increased with the use of molybdenum doped magnetite catalyst and affected mainly by the temperature and solvent: coal ratios. An oil yield of approximately 51.26% was obtained for blank runs and up to 54.77% for catalysed runs. As a hydrodesulphurisation (HDS) performer and selectivity to the production of long and branched chain alkanes, Ni-Mo had an improved performance over Co-Mo. Co-Mo is selective to a higher concentration of cycloalkanes. For 16 days on stream each, Ni-Mo had a higher activity than Co-Mo. A comparison of the actual data with a literature baseline, showed similarities for the results obtained using 2:1 solvent: coal ratios for both the blank and catalysed runs. As literature made use of severe operating conditions, the performance of the experimental batch reactor system was superior to literature. While there remains room for improvement in the design of the two-stage system, evidence exists that the potential to cover the demand for low–sulphur, crude diesel and solvents from the production of high value hydrocarbon liquid in the said process, is demonstrated.
Mercury adsorption from South African coal fired power plants using modified and unmodified fly ash.
(2024) Govender, Varnika.; Lokhat, David.; Carsky, Milan.
Mercury can be classified as a global pollutant due to its ability to travel vast distances atmospherically, be deposited on land, and enter water sources. It poses significant environmental, human, and animal risks, especially when converted to methylmercury (any toxic compound containing the complex CH3Hg) (Merriam-Webster, 2024). One of the primary anthropogenic sources is from coal-fired power stations. Therefore, effective control measures need to be employed. Fly ash provides a practical, cheap solution by acting as an adsorbent The fly ash can be enhanced by modification with metallic salts such as CuCl2, FeCl3, ZnCl2, and CuSO4. Two South African fly ash samples obtained from Camden and Majuba power stations show excellent adsorption capacity for mercury when modified by CuCl2 and FeCl3. Fly ash characterisation revealed that the two fly ash samples have very similar chemical compositions. However, the pH differs, with the Camden (dark) ash having a lower pH of 8.62 and Majuba (light) ash having a pH of 10.35. The fly ash samples successfully adsorb Hg0 whilst in the presence of CO2, which competes for the active sites available for adsorption. The emission control technology is costly, so power stations should meet a specific eligibility criterion. The fly ash adsorption process can be operated as a single pollutant system or a part of sequential pollution systems with co-benefits. The environmental impact should be an essential consideration when operating the pollution control system as the waste generated is classified as hazardous according to NEM: WA (2008).