EVALUATION OF MULTI-ZONE MODELLING FOR FIRE RISK ASSESSMENT IN TUNNELS

Abstract

Abstract
Tunnels are one of the most complex and delicate elements of the road system. Assessing fire behaviour and fire severity remains the biggest source of uncertainty in risk assessment and is an essential step in the tunnel design process. This study focuses on investigating the reliability and accuracy of different available tools to describe tunnel fire dynamics, with a focus on assessing the efficiency of the Multi-Zone (MZ) Fire model. This tool is under development thanks to a collaboration between Lund University and the Italian fire engineering company Cantene srl, with the aim of further developing a multi-zone modelling approach to be implemented in their tunnel risk analysis tool called ARTU®.
The study consists of three phases. In the initial stage a literature review was performed to investigate what type of methods and tools are available to calculate gas temperatures and smoke layer height in tunnels. In the second phase, the three selected models, CFAST, MZ and FDS, were benchmarked in set of simple tunnel configurations to investigate the influence of varying Heat Release Rate (HRR), geometry and ventilation on the results. In the last phase the tools were validated against a set of experimental data from full scale tests performed in the Koralm Tunnel in Austria, provided by Graz University of Technology.
The accuracy, the attributes, and the constraints of the tools were evaluated. The Multi-Zone (MZ) Fire model generally provided satisfying results and a good quality of agreement with more validated software. The accuracy of the tool was further discussed, recommendations for its use were provided, limitations and opportunities for further development were identified and presented in the study.

Popular Abstract

Annachiara Nardone, “Evaluation of multi-zone modelling for fire risk assessment in tunnels”
IMFSE Master Thesis, Lund University, May 2023
Popular Science Summary
Tunnels are crucial infrastructures for human activities and facilitate overpassing mountains and reducing distances, helping to reduce environmental loads and transportation costs at the same time. At the same, they are one of the most complex and delicate elements of the road system. Tragic events like the accidents in Mont Blanc Tunnel in 1999 and in St. Gotthard Tunnel in 2011 culminated in significant loss of life and substantial economic damage. For the intrinsic complexity, risk assessment is an essential step in tunnels design process. Fire severity in these cases is deeply linked to the environmental circumstances and to the infrastructure design itself. Therefore, it is very complicated to assess and predict fire behaviour.
This study focuses on investigating the reliability of several available tools to describe tunnel fire dynamics, with a focus on assessing the accuracy of the Multi-Zone (MZ) Fire model. This software is under development thanks to a collaboration between Lund University and the Italian fire engineering company Cantene srl, with the aim of further developing a modelling approach to be implemented in their tunnel risk analysis tool called ARTU®.
The study consists of three phases. In the initial stage research was conducted to find out if other studies are available and what type of methods and tools can be used to determine the main characteristics of fire in tunnels. Three tools were selected: a two-zone model (CFAST), that divides the environment into a hot upper layer and a cold bottom layer; a computational fluid dynamics model (FDS), that uses physics and fluid dynamics to describe the three-dimensional phenomena of a fire and it is very precise; a multi- zone model (MZ), that analyse the fire dividing the environment into multiple vertical and horizonal layers and represent a compromise between the other two options.
In the second phase, the three models, CFAST, MZ and FDS, were compared simulating with all of them the same set of simple tunnel configurations, to understand how different variables may affect the results. Varying fire intensity, different geometrical characteristics and the presence of mechanical ventilation were used as variables.
In the third and last phase, the tools were used to reproduced real full scale fire tests performed Koralm Tunnel in Austria by Graz University of Technology. The software results were compared to the data collected during the experiments to assess their ability to reproduce real scenarios.
The accuracy, the attributes, and the constraint of the tools were evaluated. The Multi-Zone (MZ) Fire model generally provided satisfying results and a good quality of agreement with more recognized software. The accuracy of the tool was further discussed, recommendations for its use were provided, limitations and opportunities for further development were identified and presented in the study.

• author: Nardone, Annachiara
• supervisor: Nils Johansson; Enrico Ronchi
• organization: Division of Fire Safety Engineering
• course: VBRM05 20231
• year: 2023
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• keywords: tunnels, tunnel fires, multi-zone modelling
• report number: 5699
• other publication id: LUTVDG/TVBB—5699--SE
• language: English
• id: 9118843
• date added to LUP: 2023-06-08 08:13:51
• date last changed: 2023-06-08 08:13:51