Tesi etd-02142016-195414 |
Link copiato negli appunti
Tipo di tesi
Tesi di laurea magistrale
Autore
OVIDI, FEDERICA
URN
etd-02142016-195414
Titolo
Modelling the behaviour of pressurized vessels exposed to fire with defective thermal protection systems
Dipartimento
INGEGNERIA CIVILE E INDUSTRIALE
Corso di studi
INGEGNERIA CHIMICA
Relatori
relatore Ing. Landucci, Gabriele
Parole chiave
- bleve
- defect
- defective thermal insulation
- FEM
- fire exposure
- fireball
- hazmat transportation
- LPG tanker
- lumped model
- passive fire protection
- RADMOD
Data inizio appello
04/03/2016
Consultabilità
Completa
Riassunto
Industrialized society is linked to the transport of hazardous materials by road and rail, among other. During transportation, accidents may occur and propagate among the tankers leading to severe fires, explosion or toxic dispersions. This may increase the level of individual and social risk associated to those activities, since the transport network often crosses densely populated area. The escalation of a primary event, in this case the fire, is typically denoted as domino effect, and the triggered secondary events typically are amplified.
In the framework of liquefied petroleum gas (LPG) transportation, severe fire and explosion hazards are associated to the possible catastrophic rupture of tankers, which may be induced by domino effect of accidental fires. Heat resistant coatings may protected tankers against the fire, reducing the heat load that reaches the tank shell wall and the lading. Indeed, the rupture is the result of the double effect of thermal weakening of the tank material and the increasing pressure due to LPG evaporation. However, this protection systems are not ideal and undergo defects due to both material degradation and accidental damage. Therefore, protection may be ineffective. The present work is aimed at characterizing the performance of defective coatings.
The first part of the work is devoted to the characterization of past accidents occurred in the framework of road and rail transportation of hazardous materials. The ARIA and MHIDAS databases are adopted as data sources, identifying 245 road and 220 rail accidents involving hazardous materials. The analysis highlighted the importance of protecting tank from heat load to avoid the rupture and related severe scenario. For these reasons, in North America the installation of a heat resistant coating is used to protect dangerous good tankers from accidental fire exposure. In Europe, ADR and RID regulations govern transnational transport of hazardous materials by road and by rail, respectively, and still not include any section about thermal protection systems of tankers.
Possible concerns related to the installation of these systems is due presence of defects that may be formed accidentally in the fireproofing layer. It is therefore important to establish what level of defect is acceptable in order to avoid the failure of tankers, in the prospective of a wider implementation of tankers fire protections in the European framework. Since large scale bonfire tests are expensive and difficult to be carried out in order to verify the thermal protections adopted, modelling the behaviour of pressurized insulated tankers when exposed to the fire is a possible solution to test the adequateness of defective protections.
In order to describe the thermal behaviour of real scale LPG tanks exposed to fire, a lumped model (namely, ‘RADMOD’) and a Finite Elements Model (FEM) are developed. The models are validated against available experimental data and allow predicting the thermal behaviour of tankers with defective coating when exposed to fire, with the aim to assess the thermal protection performance. The phenomena taking place through the vessel in presence of defects are investigated and characterized, in order to reproduce the experimental data on thermal behaviour of defective thermal protection systems exposed to fire.
The FEM model allows to determine the wall temperature profile and the stress distribution over the vessel, determining, in the end, a critical defect size that lead to the tank failure, with respect different fire conditions. A sensitivity analysis is performed on the FEM model in order to identify the parameters that mostly affect the heat exchanges of the system. This analysis highlights the main relevance of the flame temperature against other parameters, such as convective heat transfer coefficients and emissivity of flame and steel.
The complex analysis performed by FEM model, requires high computational times, which may be prohibitive when a wide number of runs is required. The RADMOD code is a simplified lumped model, which allows to assess the behaviour, among other, of the pressure and the fluid temperature with lower, and thus acceptable, computational time. Another plus of the RADMOD model is that it can be run for a wide range of materials, substances, geometries and fire scenario, estimating a conservative but credible time to failure of the tank. The novel mathematical code for defective thermal protection system is added to the previous version of the RADMOD model, which was implemented for unprotected or completely coated tanks, thus all the phenomena related to the defect enclosure are characterised. In addition, other phenomena, already present in the RADMOD model, are revised to enhance the potentiality of the code. The comparison of results with available experimental data on medium-scale shows that the model proposed in this thesis work can reasonably predict the thermal response. The application of the modelling tool to different geometries is performed considering real-scale defects. Thus, several case-studies were defined in order to reproduce medium- and large-scale tanks varying a few parameters, such as defect size and liquid filling level, for testing the reproducibility of the new model. The results from the case studies highlight the potentiality and the flexibility of the RADMOD code in modelling the thermal response.
The ultimate goal would be to apply the data collected from RADMOD code about temperature and pressure of lading, as boundary condition in the FEM model for an improved modelling of thermal behaviour of real-scale LPG tanks in fire scenarios even if there is a defective thermal protection system.
In the framework of liquefied petroleum gas (LPG) transportation, severe fire and explosion hazards are associated to the possible catastrophic rupture of tankers, which may be induced by domino effect of accidental fires. Heat resistant coatings may protected tankers against the fire, reducing the heat load that reaches the tank shell wall and the lading. Indeed, the rupture is the result of the double effect of thermal weakening of the tank material and the increasing pressure due to LPG evaporation. However, this protection systems are not ideal and undergo defects due to both material degradation and accidental damage. Therefore, protection may be ineffective. The present work is aimed at characterizing the performance of defective coatings.
The first part of the work is devoted to the characterization of past accidents occurred in the framework of road and rail transportation of hazardous materials. The ARIA and MHIDAS databases are adopted as data sources, identifying 245 road and 220 rail accidents involving hazardous materials. The analysis highlighted the importance of protecting tank from heat load to avoid the rupture and related severe scenario. For these reasons, in North America the installation of a heat resistant coating is used to protect dangerous good tankers from accidental fire exposure. In Europe, ADR and RID regulations govern transnational transport of hazardous materials by road and by rail, respectively, and still not include any section about thermal protection systems of tankers.
Possible concerns related to the installation of these systems is due presence of defects that may be formed accidentally in the fireproofing layer. It is therefore important to establish what level of defect is acceptable in order to avoid the failure of tankers, in the prospective of a wider implementation of tankers fire protections in the European framework. Since large scale bonfire tests are expensive and difficult to be carried out in order to verify the thermal protections adopted, modelling the behaviour of pressurized insulated tankers when exposed to the fire is a possible solution to test the adequateness of defective protections.
In order to describe the thermal behaviour of real scale LPG tanks exposed to fire, a lumped model (namely, ‘RADMOD’) and a Finite Elements Model (FEM) are developed. The models are validated against available experimental data and allow predicting the thermal behaviour of tankers with defective coating when exposed to fire, with the aim to assess the thermal protection performance. The phenomena taking place through the vessel in presence of defects are investigated and characterized, in order to reproduce the experimental data on thermal behaviour of defective thermal protection systems exposed to fire.
The FEM model allows to determine the wall temperature profile and the stress distribution over the vessel, determining, in the end, a critical defect size that lead to the tank failure, with respect different fire conditions. A sensitivity analysis is performed on the FEM model in order to identify the parameters that mostly affect the heat exchanges of the system. This analysis highlights the main relevance of the flame temperature against other parameters, such as convective heat transfer coefficients and emissivity of flame and steel.
The complex analysis performed by FEM model, requires high computational times, which may be prohibitive when a wide number of runs is required. The RADMOD code is a simplified lumped model, which allows to assess the behaviour, among other, of the pressure and the fluid temperature with lower, and thus acceptable, computational time. Another plus of the RADMOD model is that it can be run for a wide range of materials, substances, geometries and fire scenario, estimating a conservative but credible time to failure of the tank. The novel mathematical code for defective thermal protection system is added to the previous version of the RADMOD model, which was implemented for unprotected or completely coated tanks, thus all the phenomena related to the defect enclosure are characterised. In addition, other phenomena, already present in the RADMOD model, are revised to enhance the potentiality of the code. The comparison of results with available experimental data on medium-scale shows that the model proposed in this thesis work can reasonably predict the thermal response. The application of the modelling tool to different geometries is performed considering real-scale defects. Thus, several case-studies were defined in order to reproduce medium- and large-scale tanks varying a few parameters, such as defect size and liquid filling level, for testing the reproducibility of the new model. The results from the case studies highlight the potentiality and the flexibility of the RADMOD code in modelling the thermal response.
The ultimate goal would be to apply the data collected from RADMOD code about temperature and pressure of lading, as boundary condition in the FEM model for an improved modelling of thermal behaviour of real-scale LPG tanks in fire scenarios even if there is a defective thermal protection system.
File
Nome file | Dimensione |
---|---|
Tesi_Ovi...itiva.pdf | 4.43 Mb |
Contatta l’autore |