Validation of cfd codes for risk analysis of accidental hydrocarbon fires
- Autores: Francisco de Borja Rengel Darnaculleta
- Directores de la Tesis: Elsa Pastor Ferrer (dir. tes.), Eulàlia Planas Cuchi (codir. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2019
- Idioma: español
- Tribunal Calificador de la Tesis: Joaquim Casal Fàbrega (presid.), Miguel Ángel Muñoz Messineo (secret.), Micaela Demichela (voc.)
- Programa de doctorado: Programa de Doctorado en Ingeniería de Procesos Químicos por la Universidad Politécnica de Catalunya
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
Accidental releases of flammable hydrocarbons in chemical process industries can trigger severe hazards: explosions, fires, and dispersion of toxic vapour clouds. Explosions and toxic releases may injure people within large damage radius; however, fires are the most common accidental events that may lead to catastrophic consequences in terms of life and property losses.
Within this framework, the prediction of the related-fire effects may significantly contribute to identify measures needed to eliminate or mitigate the consequences of accidents in processing environments. Semi-empirical methods can provide rapid estimations of the flame-geometry descriptors as well as estimations of the heat flux received at a given distance from the fire origin. Based on that information, active protection systems and inherent safer design measures (i.e. safety distances between equipment) can be determined to prevent major fire accidents. Nevertheless, these are based on empirical and statistical data, and do not cover the overall characteristics of the fire behaviour.
Computational Fluid Dynamics (CFD) modelling can provide more detailed insights of the related fire effects considering additional complexity, such as different geometries and alternative boundary conditions, and representing different fire sizes: from small to large scale fires. Nevertheless, CFD requires detailed input data, expert knowledge on the phenomenon simulated and on the physical models implemented, and demands high computational resources. The use of CFD modelling for technological risk analysis is still incipient, so detailed validation exercises are needed before their use in real applications.
This thesis is mainly aimed at assessing the predictive capabilities of different CFD codes (FDS, FLACS-Fire and FireFOAM) when predicting the hazardous effects of hydrocarbon pool fires and jet fires. Specifically, large-scale pool fires of diesel and gasoline (from 1.5 to 6 m-diameter), vertical sonic jet fires of propane (from 0.09 to 0.34 kg/s with orifice diameters of from 10 to 25.5 mm), vertical subsonic jet fires of methane in normal- and sub- atmospheric pressures (from 0.6 to 1 bar with an orifice diameter of 3 mm), and vertical and horizontal subsonic jet fires of propane (from 0.007 to 0.11 kg/s with orifice diameters of from 12.75 to 43.1 mm-diameter) have been modelled in different CFD codes. Prescribing burning rates provide accurate predictions of the pool fire effects with maximum cell sizes of 0.2 m. On the other hand, the cell sizes of sonic and subsonic jet fires should be determined by considering a fire characteristic diameter of 16 and 12, respectively. A minimum number of 400 solid angles is recommended to obtain accurate estimations of the thermal flux.
Based on the numerous computational simulations performed, Best Practice Guidelines (BPG) are developed to determine a code as ‘valid’ or not, and to provide guidance on the most suitable modelling settings when performing CFD simulations of accidental hydrocarbon fires. The BPG usefulness is proved through a case study of an oil storage farm located in the Port of Barcelona. Large over-estimations of the heat flux values are found with semi-empirical correlations and thus, the safety measures required would be very conservative and costly. Therefore, CFD modelling is recommended method to perform detailed FHA in chemical and process industries.
