The Refcoal process is being developed to produce graphite from coal. Coal is dissolved in dimethylformamide (DMF) and sodium hydroxide (NaOH) is used as additive. After separation, the extracted coal (Refcoal) is precipitated with water and dried. The extraction process and subsequent solid-liquid separation processes have to be as efficient and cost-effective as possible. The purpose of the study was therefore to complete research on various unresolved aspects of the processes as identified by the candidate and supervisor. Extraction at 95 °C (DMF:coal:NaOH = 100:10:1), has an induction period of approximately 60 minutes observed, after which the reaction rate increases considerably. The reaction reaches completion after 360 minutes. An increase in stirring rate decreases extraction time due to the elimination of external mass-transfer limitations. The progress curves obtained for extraction at 135 °C with lower solvent-to-coal ratios differ dramatically from those obtained in previous studies, which indicates that changes in the raw materials and the experimental set-up have a great influence on the extraction at higher temperatures and concentrations. These extractions at higher temperatures using DMF:coal:NaOH ratios between 100:30:3 and 100:30:2 take approximately 360 minutes to complete and do not have an induction period as is the case with the extractions at 95 °C. It was found that the optimum DMF:coal ratio for an operating temperature of 135 °C, is 10:3. The high-temperature extractions reach completion in different time periods, depending on the amount of NaOH added to the reaction mixture. When very low concentrations of NaOH are added, the extraction will take much longer to complete and vice versa. The amount of NaOH used influences various aspects of the process. The cost analysis of the process falls beyond the scope of this investigation, but it is recommended that a thorough financial study is done to determine the optimum balance between raw materials, heat load and plant availability. The relationships between the concentration of Refcoal in the Refcoal solution and the absorbance values measured are polynomial expressions ending in downward concaves. The kinetics for the low-concentration (DMF:coal:NaOH = 100:10:1) extraction are best described by an autocatalytic reaction rate equation which is a function of coal, coal complex and NaOH concentration. A good fit was also obtained for the high temperature extractions. The rate expression is a function of both the coal and NaOH concentrations, but not of the coal complex. The sedimentation test showed promising results. The use of a thickener instead of a centrifuge to separate the insoluble material from the Refcoal solution would be a feasible cost-saving method. Filtration of the Refcoal solution (after centrifugation) using suitable filter media decreases the amount of impurities in the Refcoal. Filtration constants were determined for the best filter medium. The use of a hydrocyclone to separate the insoluble material from the extract is not recommended as it did not give the required efficiency to make the process viable. It is recommended that more tests be done under different conditions. Useful expressions were obtained for the change in viscosity with temperature for three different concentrations of Refcoal solution. It was determined that the viscosity of the Refcoal solution increases with time and it is therefore recommended that this be taken into account when equipment is being designed and plant scheduling is being done. / Dissertation (MEng (Chemical Engineering))--University of Pretoria, 2007. / Chemical Engineering / MEng / unrestricted

A new process is being developed to produce graphite from prime coking coal. Coal is dissolved in dimethylformamide (DMF), on addition of sodium hydroxide. The minerals and undissolved coal are separated by centrifugation and filtration to give a solution (referred to as Refcoal solution or RCS). Over 90 wt % of the organic part of a flotation product, from the Tshikondeni mine, can be dissolved at temperatures ranging from room temperature to 135°C. The purified coal (referred to as Refcoal) and DMF need to be separated. the Refcoal to be coked and the DMF to be purified and recycled. This process should be as efficient as possible, whilst both products should be low in water content to minimise drying costs. The addition of water to the Refcoal solution causes precipitation to take place, forming a gel (referred to as Refcoal gel) liquid system. This mixture can be either centrifuged or filtered to give a denser gel, containing water, DMF and coal solids, and supernatant or filtrate, containing water and DMF. Different techniques and processes can be used to improve the separation of the DMF from the Refcoal by achieving a denser Refcoal gel: • Longer centrifugation times improve the density and therefore the separation, but this technique has its limits. • The use of low-temperature water improves the separation. • The use of syneresis could improve separation at a lower cost: heated standing tanks are used to expel the supernatant and therefore increase the density of the gel, thereby decreasing the required number of washing stages. • The addition of toluene at the beginning of a wash improved the removal of DMF by 20%, using centrifugation as separation method. • Pressure filtration gave a 20% improvement on centrifugation, with no additives. • The addition of toluene to the pressure filtration process gave another improvement of 15%, and after three stages the percentage of solids in the gel was 28%, the highest so far achieved. This method also resulted in the highest removal of DMF in the first stage (73% of the original DMF in the RCS was removed). Counter-current washing shows the greatest potential, using the least amount of water. The concentration of DMF in the wash solution, to gel the Refcoal solution, is a limitation of this process. If the concentration is too high, no gelling and therefore no separation can take place in the first stage. It is recommended that counter-current washing using pressure filtration should be investigated; however, this will be difficult on a laboratory scale due to the mass losses during transfers. / Dissertation (MEng (Chemical Engineering))--University of Pretoria, 2009. / Chemical Engineering / unrestricted