An Assessment of the Spread Rates of Accidentally Spilled Flammable Liquids and the Development of a Model to Forecast the Spill Velocity

摘要

More often than not, major fires and explosions occur in chemical process industries as also in public places, due to accidental spills of flammable substances. This makes it important to study the dynamics of flammable liquids spilled onto the ground, so that strategies to contain and quench such spills can be devised. Moreover, the risk posed by a liquid spilling over an inclined plane is much greater than the risk posed by a spill over a flat surface because in the former case there is an increased rate of evaporation due to an enhanced spread rate, thereby enhancing the possibility of the formation of highly flammable/explosive vapor clouds. Such vapor clouds are known to drift to other units and damage them, often escalating a mere spill into a major catastrophe. But even though a large number of accidents have occurred where the flammable liquid had spilled over inclined planes, little work has been done toward understanding the dynamics of such spills which may help in the development of models to forecast the pattern and impact of such spills. The present work is an attempt to fill the knowledge gap and is arguably the first-ever study of its kind. In this work, the dynamics of spill of three different fuels-petrol, kerosene and ethanol-on two types of substrates of different roughness-vinyl and plastered concrete-have been studied. The angle of inclinations over which the spill dynamics have been studied is 0 degrees, 5 degrees, 10 degrees, 15 degrees and 20 degrees to the horizontal. Further, for the first time, the application of artificial neural networks to predict the spill velocity has been demonstrated.