Fire Safety System Effectiveness for a Risk-Informed Design Tool
Thesis DisciplineCivil Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
The purpose of this research is to identify how uncertainty in fire safety system effectiveness should be considered in a new risk-informed design fire tool, B-RISK. Specific objectives were to collect the available data on fire safety system effectiveness from the literature, investigate methods to improve fire safety system effectiveness data collection, develop the risk-informed design fire tool to propagate the uncertainties, and recommend methods to rank the sources of uncertainty for fire safety system effectiveness for appropriate model selection. The scope of the research is limited to the effects of systems on fire development and smoke spread and does not include the effects of the fire on systems (such as loss of structural integrity) or interactions with occupants. Sprinkler effectiveness data from recent New Zealand Fire Service data is included with a discussion of the uncertainty in this type of data and recommendations for improving data collection. The ability of the model to predict multiple sprinkler activations is developed in conjunction with a hydraulic submodel in B-RISK to include water supply pressure effects on sprinkler effectiveness. A new method of collecting reliability data on passive fire protection elements such as doors was developed. Data collected on the probability for doors in shared means of escape to be open and the time doors are open during occupant evacuation using this method is presented. Available data on smoke management system effectiveness is listed, along with a discussion of why there is more uncertainty associated with these systems compared with sprinkler systems. The capabilities of B-RISK for considering fire safety system effectiveness are demonstrated using Australasian case studies.