Validation and Improvement of Phenomenological Model for MEC Prediction Through Experimental Investigation on Dust Vaporization and MEC: Application to Sulfur Dust

Abstract

The amount of published research on the MEC of sulfur dust remains rather limited. This work aims to contribute to the development of an existing phenomenological model to predict the Minimum Explosion Concentration (MEC) by eliminating its limitations. The limitations of the model studied lie within the vaporization sub-model, ignition source temperature, and particle size definition. A Thermogravimetric analyzer was used to measure the vaporization rate of sulfur. The flux increased with temperature but an unusual decrease between 300°C and 350°C was observed. This is due to the dissociation energy of S8 and S5 to S2 which is consumed from the vaporization energy supplied. The vaporization sub-model fit the experimental data up to 300°C but was incapable of fitting the rest because it assumes all the heat supplied is used to vaporize the sample. The MEC of five sieved sulfur dust samples was measured using a modified Hartmann tube with a heated coil. The MEC increased with particle size and decreased with ignition source temperature which conforms with the literature. The model was modified by changing the: ignition source temperature, time, and molecular weight. Altwal’s assumed ignition source temperatures were replaced with extrapolated temperatures. The molecular weight was adjusted accordingly (solid: 256.65 g/mol, vapor: average molecular weight). This overestimated the MEC, due to the formation of smaller species which reduce the number of mols produced in the vapor, thus requiring a higher concentration to reach LFL. The ignition time correlation was estimated using MEC measurements of this study and it showed a better dependence on the ignition source temperature. The modifications above were all applied to the model but was not capable of predicting the MEC due to limitations in LFL modeling. The Weibull distribution was proposed to express the MEC as a function of the entire particle size distribution using a shape and scale parameter. The results show a surface plot with a clear trend on the influence of both parameters on the MEC. Increasing the parameters increases the particle size of the sample resulting in a higher MEC.