No description available.

Over the past years, catastrophic dust explosion incidents have caused numerous injuries, fatalities and economical losses. Dust explosions are rapid exothermic reactions that take place when a combustible dust is mixed with air in the presence of an ignition source within a confined space. A variety of strategies are currently available to prevent dust explosion accidents. However, the recurrence of these tragic events confirms flaws in process safety for dust handling industries. This dissertation reports advances in different approaches that can be followed to prevent and mitigate dust explosions. For this research, a 36 L dust explosion vessel was designed, assembled and automated to perform controlled dust explosion experiments. First, we explored the effect of size polydispersity on the evolution of aluminum dust explosions. By modifying systematically the span of the particle size distribution we demonstrated the dramatic effect of polydispersity on the initiation and propagation of aluminum dust explosions. A semi-empirical combustion model was used to quantify the laminar burning velocity at varying particle size. Moreover, correlations between ignition sensitivity and rate of pressure rise with polydispersity were developed. Second, we analyzed the effect of particle size and crystalline levels in the decomposition reactions of explosion inhibitor agents (i.e., phosphates). We fractionated ammonium phosphate- monobasic (NH_4H_2PO_4) and dibasic ((NH_4)_2HPO_4) at different size ranges, and synthesized zirconium phosphate (Zr(HPO_4)_2·H_2O) at varying size and crystalline levels. Particle size was found to be crucial to improve the rate of heat absorption of each inhibitor. A simplified model was developed to identify factors dominating the efficiency of dust explosion inhibitors. Finally, we conducted computational fluid dynamic (CFD) simulations to predict overpressures in dust explosions vented through ducts in large scale scenarios. We particularly focused on the adverse effects caused by flow restrictions in vent ducts. Critical parameters, including ignition position, geometric configuration of the vent duct, and obstructions of outflow such as bends and panels were investigated. Comparison between simulation and experimental results elucidated potential improvements in available guidelines. The theoretical analyses complemented the experimental work to provide a better understanding of the effects of particle size on the evolution of dust explosions. Furthermore, the validation of advanced simulation tools is considered crucial to overcome current limitations in predicting dust explosions in large scale scenarios.

Depuis plusieurs années déjà, nombre d’applications industrielles impliquant des nanomatériaux ont vu le jour mais les connaissances relatives aux dangers de ces nouveaux matériaux sont actuellement assez restreintes. Le développement de ces nouveaux produits ne pouvant se poursuivre sans une évaluation approfondie des risques pour l’environnement et au poste de travail, les dangers relatifs aux nanoparticules doivent être évalués. La toxicité potentielle de ces nouveaux matériaux est souvent mise en avant. Néanmoins, les risques d’incendie et d’explosion ne doivent pas être négligés. Centrées essentiellement sur les poudres de taille micrométrique, les données de la littérature ne permettent pas, en effet, à l’heure actuelle, d’évaluer la probabilité et la gravité d’une explosion de nanopoudres. Dans ce contexte, la sensibilité à l’inflammation et la sévérité d’explosion de nanomatériaux pulvérulents typiques ont été évaluées ainsi que la validité des appareillages et procédures standards, habituellement utilisés lors d’une telle démarche. Enfin, la méthodologie adoptée afin d’évaluer les risques d’inflammation et d’explosion d’une installation de production de nanopoudres et de sécuriser au mieux la santé des travailleurs exposés aux nanoparticules est illustrée aux travers de deux exemples. Cette démarche pourra servir de base à de futures analyses de risques concernant les produits nanostructurés, exercice qui va devenir indispensable et de plus en plus fréquent au vu du contexte économique et réglementaire / In the industrial and research fields, nanomaterials provides a growing interest and many industrial applications have already been developed in the last years. However, knowledge about the hazards related to these new materials is currently limited. As safe nanomaterial production cannot be permitted without a deeper evaluation of environmental and occupational hazards, hazards related to nanoparticles have to be evaluated. One often thinks about the potential toxicity of nanoparticles. However, dust fire and explosion should not be neglected when the dusts are combustible, which may often be the case. So far, literature studies concerning the evaluation of explosion and flammability risks of powders were essentially carried out on micron-sized materials and do not enable in fact to evaluate fire and explosion risk probabilities and gravities of nanopowders. The main goal of this work is to study explosion and ignition risks related to nanopowders. In particular, the evaluation of the explosion sensitivity and severity of typical nanomaterials has been studied as well as the validity of the existing analytical and methodological tools designed to evaluate dust ignition and explosion hazards. This work also deals with the methodology applied to a plant and to a laboratory in order to define the best safety barriers which were positioned to ensure the best occupational safety level to all workers and evaluate in a good way the ignition and explosion risks related to the use and production of fluffy nanomaterials. This work will certainly help risk engineers concerned about the handling and the production of combustible nanopowders.

<p></p><p>In the grain handling and processing industry, dust emission and accumulation are major concerns for the safety of workers and for explosion risks. Dust emission and accumulation locations highly depend on the facility design and equipment used for handling and processing. To prevent an explosive atmosphere, monitoring the amount of dust accumulated or dispersed is extremely important. However, methods of measuring the dust concentration require the installation of equipment. The Occupational Safety and Health Administration (OSHA) regulations and National Fire Protection Association (NFPA) standards restrict the thickness of dust layers on floors for fine powder materials such as starch. The objective of this dissertation was to better understand the rate of dust layer accumulation, dust suspension patterns, and the optical properties of suspended dust. For this purpose, The Discrete Phase Model (DPM) was combined with a Computational Fluid Dynamics Model (CFD) and the hybrid model was used to model dust dispersion. Dust dispersion patterns under pressure, such as primary explosions or leakage from equipment, were simulated using the unsteady CFD-DPM approach. The particle-wall interaction based on energy conservation was also introduced in this model. Both one-time and continuous dust dispersion in an enclosed chamber were simulated to mimic secondary explosions and the dust emission from processing equipment. In addition, the light extinction property of suspended dust was studied as a method of measuring suspended dust concentration. </p> <p>For a one-time dust dispersion incident, the predicted dust concentration agreed with the simulation result for the trial conducted at a dust injection velocity of 2 m/s with injection rates of 0.05 and 0.10 kg/m³ and at a dust injection velocity of 10 m/s with an injection rate of 0.05 kg/m³. The dust concentration in the entire chamber increased with dust injection velocity and the mass of injected dust. As dust injection velocity increased, dust spread out and formed a larger explosive dust cloud. However, the dust concentration inside the chamber was non-uniform. Considering the minimum explosive concentration, the largest explosive cloud was created at a dust injection velocity of 10 m/s with an injection rate of 0.10 kg/m³. Explosive concentrations of dust were found somewhere in the chamber for all dispersion rates. At an injection velocity of 10 m/s with an injection rate of 0.10 kg/m³, the predicted dust concentration was 10% more than the measured dust concentration. Thus, this model is suitable for dilute dust particle dispersion flows, where the volume fraction of particles is less and only a single particle layer settles.</p> <p>Continuous dispersion was simulated to determine the suspended dust concentration and particle deposition patterns. Dust was dispersed for 30 s at dispersion rates of 2, 4 and 6 g/min at a dust injection velocity of 2 m/s. The dust concentration increased at a constant rate after a few seconds of dispersion, regardless of the dust dispersion rate. Most dust particles were deposited near the dust dispersion nozzle. Large particles were more affected by gravitational force and inertia compared with small particles, which traveled with airflow and settled behind the nozzle. The dust accumulated close to the dispersion nozzle faster than behind the nozzle location. However, specific attention must be paid to small particles, because they are more likely to cause an explosion, as their minimum explosive concentration is lower than that of large particles.</p> <p>The light extinction coefficients of cornstarch, grain dust, and sawdust were measured using a two-target method. The suspended dust concentration was measured using a calibrated laser instrument. The light extinction coefficient was linearly related to the suspended dust concentration. The correlation coefficient between the light extinction coefficient and suspended dust concentration depended on particle size, particle shape, and chemical properties. </p> <p>Controlling dust cloud generation and minimizing the concentration and volume of dust clouds are some key measures to prevent dust explosions. The mathematical models developed in this study to predict dust dispersion, suspension, and rate of settling will help solve a few of the challenges in the particulate material handling and processing industry. This method of measuring the light extinction coefficient can be applied development of a dust safety monitoring system. The result presented in this dissertation will help the industry prevent the formation of an explosive atmosphere.</p><br><p></p>

La sensibilité et la sévérité d'explosion des différents mélanges gaz/vapeur-poussière ont été étudiées grâce à des dispositifs standards (sphère de 20 L, tube de Hartmann). Les spécificités des explosions de mélanges hybrides gaz/poussière ont été mises en évidence. En fait, même pour des concentrations de gaz inférieures à la limite inférieure d'explosivité (LIE), la probabilité d'inflammation et la gravité d'explosion peuvent être considérablement augmentées, ce qui permettra notamment de conduire à de grands changements dans la détermination des zones ATEX. Il a été, par exemple, démontré que ces mélanges peuvent être explosifs même lorsque la concentration en poudre et la concentration en vapeur sont respectivement en dessous de la concentration minimale explosive et de la LIE. En outre, des effets de synergie ont été observés et la vitesse de montée en pression de mélanges hybrides peut être supérieure à celles des gaz purs. Les origines de ces spécificités ne doivent pas être recherchées dans la modification d'un paramètre unique, mais peuvent probablement être attribuées aux effets combinés sur l'hydrodynamique (propagation de la flamme), le transfert thermique et la cinétique de combustion. Des expériences ont été menées afin de souligner l'importance de chaque contribution. Basé sur des schémas cinétiques classiques à coeur rétrécissant prenant en compte des diverses contraintes lors d'une réaction non-catalytique de gaz/solide et sur des modèles de combustion homogène pour les gaz, un modèle a été développé pour représenter l'évolution temporelle de la pression d'explosion pour ces mélanges / The explosion sensitivity and severity of various gas/vapor-dust mixtures have been studied thanks to specifically modified apparatuses based on a 20 L sphere and a Hartmann tube. The specificities of gas/dust hybrid mixtures explosions have been highlighted. In fact, even for gas concentrations lower than the lower explosivity limit (LEL), the ignition probability and the explosion severity can be greatly increased, which will notably lead to great changes in the Ex zones determination. For instance, it has been shown that such mixtures can be explosive when both the dust and gas concentrations are below their respective minimum explosive concentration and LEL. Moreover, synergistic effects have been observed and the rate of pressure rise of hybrid mixtures can be greater than those of the pure gases themselves. The origins of these specificities should not be sought in the modification of a single parameter, but could probably be attributed to combined impacts on hydrodynamics (flame propagation), thermal transfer and combustion kinetics. Experiments have been carried out in order to underline the significance of each contribution. Based on classical shrinking core models taking into account the various limitations during a non-catalytic gas/solid reaction and on homogeneous combustion for gases, a model has been developed to represent the time evolution of the explosion pressure for such mixtures

’explosion est un phénomène redouté dans les installations industrielles. Le risque d’explosions impliquant des poussières combustibles est présent dans un grand nombre d’industries d’activités différentes, compte tenu de la grande diversité de poussières combustibles : les poussières organiques (farine, charbon, sucre…) mais aussi les poussières métalliques (aluminium, magnésium…). En effet, toutes ces poussières combustibles si elles sont suffisamment fines, et si elles sont en suspension dans l’air, peuvent provoquer des explosions. Les industriels doivent donc quantifier et maîtriser ce risque au sein de leurs différentes installations. Dans le cas des explosions de gaz, l’état actuel des connaissances permet une compréhension et une modélisation précise du phénomène. Cependant, l’état des connaissances est plus limité dans le cas des explosions de poussières, notamment à cause de la plus grande difficulté à étudier ces dernières expérimentalement. Des modèles, basés sur les explosions de gaz, existent néanmoins dans le cas des explosions de poussières. Ces derniers semblent cohérents dans le cas d’explosions de poussières organiques mais inadaptés au cas des poussières métalliques.Ces travaux de thèse s’intéressent à l’étude expérimentale de la propagation de la flamme lors d’une explosion de poussières d’aluminium. Afin de modéliser une propagation éventuelle de flamme lors d’une explosion, une première approche expérimentale est nécessaire. Pour cette étude expérimentale des prototypes ont été spécialement conçus, puis améliorés, au cours des différents tests réalisés. La description de ces travaux est divisée en deux parties.Dans un premier temps, la mise en suspension de la poudre est étudiée. En effet, afin de pouvoir étudier ce phénomène d’explosion, un système de mise en suspension de la poudre a été élaboré. Une première partie de l’étude permet donc de s’assurer que la suspension obtenue est homogène en termes de concentration. Par la suite, le niveau de turbulence obtenue dans l’enceinte après la fin de la mise en suspension de la poudre est étudié. En effet, ce paramètre influe grandement sur la propagation de la flamme, augmentant ainsi les conséquences de l’explosion.Par la suite, la propagation de la flamme est étudiée. Pour cela, la suspension précédemment obtenue est enflammée à l’aide d’un arc électrique. Le phénomène est étudié au travers de la visualisation de la propagation de la flamme et par l’évolution de la pression dans le prototype. Deux principales méthodes optiques, l’une basée sur la visualisation de la lumière émise par la flamme et l’autre sur la visualisation de variations d’indice de réfraction (liées à des variations de température), sont utilisées. A partir de ces dernières la vitesse de propagation de la flamme dans le référentiel du laboratoire est étudiée. Cependant, cette vitesse dépend fortement du prototype utilisé pour son étude. Ainsi, une méthode est utilisée afin d’en déduire la vitesse de combustion, correspondant à la vitesse de consommation des réactifs par la flamme. Des limites potentielles de cette méthode sont par la suite exposées, et une nouvelle méthode de détermination de cette vitesse est alors proposée. / Explosions are one of the most feared events in the industry. Risk of explosions with combustible dusts can occur in a large variety of industry of different fields, because of the large amount of combustible dusts: organic dusts (flour, carbon, sugar…) but also metallic dusts (aluminum, magnesium…). All of these combustible dusts, if they are fine enough, and if they are dispersed in the air, can cause explosions. Companies have to quantify this risk present in their plant. Concerning gas explosions, the current state of knowledge allows an understanding and a precise modelling of the phenomenon. However, the state of knowledge about dust explosions is more limited, especially because of the difficulty to study the explosions experimentally. Some models, based on gas explosions, exist for the case of dust explosion. These models seem coherent in the case of organic dust explosions but less adapted for metallic dust.This PhD work focus on the experimental study of flame propagation during an aluminum dust explosion. To model an eventual propagation of the flame during the explosion, an experimental approach is required. For this experimental study, specific prototypes have been elaborated, and then improved, during the different tests. This work is mainly separated in two parts.In a first part the dispersion of the dust is studied. Indeed, to study the explosion phenomenon, a system has been elaborated to disperse the dust. A first part of study allows checking that the dispersion is well homogeneous in terms of concentration. Then, the turbulence level inside the prototype after the end of the dispersion is studied. Indeed, this parameters influence a lot the flame propagation, increasing the consequences of the explosion.Then, the flame propagation is studied. The dust dispersion, previously studied, is ignited by an electric spark. The phenomenon is studied thanks to visualization of the flame propagation and by the evolution of the pressure inside the prototype. Two main optical techniques, one based on the light emitted by the flame, the other one linked to refractive index variations (due to temperature variations) are used. Thanks to these methods, the propagation velocity in the laboratory referential is studied. However, this velocity depends mainly on the prototype used for his determination. A method is used to determine the burning velocity (consumption rate of the reactants by the flame front). Some potential limits of this method are then exposed, and a new method of determination of this burning velocity is proposed.

The current research is aimed at investigating the explosion behaviour of hazardous materials in relation to particle size. The materials of study are titanium powders having size distributions in both the micron- and nano-size ranges with nominal size distributions: -100 mesh, -325 mesh, ?20 ?m, 150 nm, 60-80 nm, and 40-60 nm. The explosibility parameters investigated explosion severity and explosion likelihood for both size ranges of titanium. Tests include, maximum explosion pressure (Pmax), maximum rate of pressure rise ((dP/dt)max), minimum explosible concentration (MEC), minimum ignition energy (MIE), minimum ignition temperature (MIT) and dust inerting using nano-titanium dioxide. ASTM protocols were followed using standard dust explosibility test equipment (Siwek 20-L explosion chamber, MIKE 3 apparatus, and BAM oven). The explosion behaviour of the micron-size titanium has been characterized to provide a baseline study for the nano-size testing, however, nano-titanium dust explosion research presented major experimental challenges using the 20-L explosion chamber.