Experimental and analytical investigation of pressurized vessels exposed to fire

The possible occurrence of accidental fires impacting vessels for transportation and storage of hazardous materials represents a key safety issue in the process industry. In these situations, the vessel heats up, pressurizes and can fail catastrophically, generating devastating consequences. Such scenarios have been extensively investigated in the past decades to improve vessel design and emergency response planning. Numerous field studies and laboratory scale tests were carried out and several models were developed to predict the thermal and mechanical response of vessels exposed to fire. However, previous modeling approaches suffer several limitations and need to be improved and experimental data is not sufficient to effectively support the development of advanced modelling tools such as CFD. With the aim of overcoming these limitations, a novel research program was proposed. This combines fire tests, carried out by means of an innovative experimental apparatus, and a CFD based modelling approach. The present work focuses mainly on the modelling part (the experimental setup is briefly described and preliminary analysis of the data from tests is presented). Starting from previous approaches presented in literature, an improved CFD modelling setup was developed. Conditions of several fire tests involving LPG and water tanks were simulated and the results are compared with experimental measurements highlighting strengths and limitations of the modelling. In the last part of the work, an alternative approach is presented, based on models developed for the study of subcooled boiling flows that showed promising results in other fields of application. The aim was to explore the possibility of extending this approach to the case of fired vessels. The work proves that CFD is a powerful tool for the development of models able to accurately describe and predict the response of a pressure vessel exposed to fire. However, further work is needed especially regarding submodels for boiling.