Assessment of the Explosion Characteristics of Dust Clouds: Standards Versus Reality

The dust explosions are a matter of main interest in the process safety. These events might occur when a combustible dust is dispersed near an ignition source in a confined facility. This fact implies that the generation of a dust cloud represents a hazard for industries that store or handle this type of combustible materials. For this reason, some organic, plastic and metal powders, which are usually considered as non-hazardous materials, have led to several incidents that were associated to explosive atmospheres.

For the characterization of a combustible dust, some specific tests have been standardized in order to determine its explosivity parameters. For instance, the 20L sphere allows the determination of the minimum explosive concentration (MEC), the maximum explosion pressure (P_max), the maximum rate of pressure rise (dP/dt)_max and the deflagration index (Kst). One of the most common assumptions associated to the experimental results of this test is based on the homogeneity of the dust cloud inside the 20 L dispersion chamber. However, due notably to the particle size distribution and the dispersion nozzle geometry, this assumption is far from being validated experimentally. With regard to the strong influence of the heterogeneity and turbulence level with the dust cloud on the flame propagation, a descriptive study has been focused on the characterization of the behavior of the dust cloud and its influence on the determination of the explosion characteristics.

The results suggest that the observed variability of the experimental data can be associated to variations of the particle size distribution of the dust and the turbulence levels before the dust ignition. Further analyses have also established that the phenomena of agglomeration and fragmentation of some dusts are evidenced inside the sphere and are related to the internal conditions of the dust dispersion.

Therefore, this paper presents a descriptive analysis of the dispersion of micrometric wheat starch within the 20 L sphere. For this purpose, this study integrates some computational tools based on the Computational Fluid Dynamics (CFD) and Discrete Element Methods (DEM). Additionally, this analysis provides a characterization of the two-phase flow inside the vessel by implementing empirical and numerical models that define the cohesive behavior of the combustible dust and the turbulence levels of the dust cloud. Furthermore, the description of the dispersion process has been enhanced by performing experimental tests based on Particle Image Velocimetry (PIV) and in situ particle size measurements. Moreover, the results have shown the inherent relation between the operating variables of the test and the variations in the size distribution of the dust. This fact has allowed considering some modifications in certain operating parameters, such as the design of the dispersion nozzle, the ignition delay and the injection pressure, in order to vary the turbulence levels of the two-phase flow.

Finally, this study establishes the high relevance of the characterization of the solid material prior to testing and also proposes some improvements in the explosion tests that will improve the reliability of the flammability parameters of the combustible dust.