Abstract
Since the beginning of modern industry, the need to move large quantity of raw materials and products has increased continuously, matching with the improvement of the transport routes, especially public roads. With more and more trucks on world’s streets, the number of accidents has increased together with the risk connected to side events derived from the transport of hazardous materials such as fuels, acids, toxic wastes and so on. In a world where there is an increasing attention to safety, transport of hazardous materials has become one important aspect to evaluate. In order to reduce the risk, a proper choice of the transport route and a reliable simulation of the consequences of possible accidents is needed. To estimate the consequences of an accident involving flammable materials, mathematical models developed in the frame of the Computational Fluid Dynamics (CFD) are able to provide a detailed description of temperatures and species concentrations but, since they are very demanding in terms of both CPU-time and analyst skill, they are not often used for a screening evaluation of the magnitude of the consequences of an accident (as required by an ordinary shipping operation). Aim of this work is to develop a simplified mathematical model able to evaluate the effects of an accident involving the spill of a fuel (in this case study, gasoline) from a tanker in a road tunnel with a subsequent fire. The simplified model should be able to produce results comparable with those obtained using CFD models with much less CPU-time, in order to be easily implemented in a risk analysis software. To achieve this purpose, a set of differential equations has been written to describe the main phenomena involved in a tunnel fire accident; moreover, all space-time correlations have been sundered to obtain a set of equations easily solvable, therefore avoiding the time-consuming operations due to multiple numerical resolutions. The results obtained for the investigated case study using the simplified model are in fairly good agreement with those obtained from a dedicated CFD model.
