Kinetic modelling of scrap tyre pyrolysis and oxidative desulphurisation of tyre-derived oil

Abstract

The amount of tyres generated around the world has been on the rise. This has prompted the need to explore ways in which waste tyres can be disposed. One of the ways of recycling waste tyres is through pyrolysis, a process that has the potential to produce oil products with high calorific value. However, the oil produced from waste tyre pyrolysis has a high sulphur content, resulting in high levels of toxic emissions during combustion. This research study focuses on two aspects associated with waste tyre pyrolysis. The first aspect deals with the establishment of the kinetics of scrap tyre pyrolysis while the second aspect involves the oxidative desulphurisation (ODS) of tyre-derived oil (TDO). The ODS was carried out with incorporation of the central composite design (CCD) technique of the response surface methodology (RSM) in order to model the desulphurisation process. In order to study the kinetics of the scrap tyre pyrolysis, three different models were applied to thermogravimetric data. The thermogravimetric (TG) experiments were carried out in a nitrogen environment and a temperature range of 20°C to 600°C at heating rates of 2, 5, 10, and 20 °C min-1. The models used to determine the activation energy (Ea) were Kissinger-Akahira-Sunose (KAS), Flynn-Wall-Ozawa (FWO) and Friedman (FR) whereas the Coats-Redfern (CR) model aided in the determination of the pre-exponential factor. The FWO model had the highest average value (R2 = 0.9847) of the coefficient of determination, and therefore the Ea values from this model were loaded into the CR model to determine the pre-exponential factors and the order of the reaction model. The thermal decomposition started at a mean temperature of about 285°C and was complete at about 482 °C for the four heating rates. Results indicate that the mass losses become greater with increasing heating rates. The thermogravimetric analysis results revealed that tyre pyrolysis involves three stages i.e. removal of lubricants and oil in the scrap tyre, breakdown of natural rubber and breakdown of butadiene rubber and styrene-butadiene rubber. The average activation energies obtained were 206.01 kJ mol-1, 206.08 kJ mol-1 and 204.82 kJ mol-1 using KAS, FWO and FR models respectively. A mean pre-exponential factor of 1.27E+10 min-1 was obtained. In addition, the results showed that the pyrolysis of the tyre crumb conforms to the second order reaction model (F2). The second objective of the study involved the use of the CCD methodology to investigate the interaction of parameters during the ODS of tyre-derived oil. The oxidative desulphurisation involved the investigation of the interaction of formic acid and hydrogen peroxide amounts, reaction time and temperature. The liquid-liquid extraction was carried out using two different solvents. In addition, for one of the solvents (acetonitrile), two solvent to oil ratios were used, bringing the total number of solvent extraction scenarios to three. The three extraction scenarios were acetonitrile to oil ratio of 1:1, acetonitrile to oil ratio of 1:2 and dimethylformamide to oil ratio of 1:1. The ODS for each of the solvent extraction scenarios consisted of 21 experiments. Therefore, the total number of experiments for the three solvent extraction scenarios was 84 i.e. 21 runs for the oxidation stage and 63 runs for the solvent extraction. The maximum sulphur removal achieved was 86.05, 52.77 and 35.00 % respectively for oxidised oils extracted using 1:1 acetonitrile to oil ratio, 1:2 acetonitrile to oil ratio and 1:1 dimethylformamide to oil ratio while the corresponding minimum sulphur removal values were 34.02, 27.91 and 3.8 %. The results of the sulphur removal in which extraction was carried out at 1:1 acetonitrile to oil ratio were further analysed and modelled. From the analysis of variance (ANOVA), the reduced cubic model was found to be the best predictor of sulphur removal during the ODS process. Coefficient of determination (R2 = 0.9776), adjusted R2 = 0.9254, predicted R2 = 0.8356 all showed that the model was significant. Moreover, the p-value for the lack of fit was 0.8926, which is indication of its insignificance relative to pure error. In summary, the data obtained from the kinetic study of the scrap tyre pyrolysis could play an important role in the design and optimisation of industrial scale scrap tyre pyrolysis units. The findings could provide an insight for improvement of the general operability of scrap tyre thermal conversion processes via pyrolysis. Further research should be carried out to obtain thermogravimetric data at higher heating rates, which can then be used to model the process via the non-isothermal means. On the other hand, the findings from the oxidative desulphurisation of tyre-derived oil obtained from this research can play a key role in minimising the levels of emission during the combustion of oils obtained from the pyrolysis of waste tyres. In addition, the knowledge from the present study could be useful in commercialisation of efficient sulphur removal technology in tyre-derived oils, which can then be blended with other fuels such as diesel for use in real combustion processes. Further work with incorporation of a different organic acid, such as acetic acid and hydrogen peroxide as part of the oxidation system may be carried out to investigate the change in the extent of sulphur removal.