Characterising a Design Fire for a Deliberately Lit Fire Scenario (2008)
Type of ContentTheses / Dissertations
Thesis DisciplineFire Engineering
Degree NameMaster of Engineering in Fire Engineering
PublisherUniversity of Canterbury. Civil Engineering
AuthorsRichards, Paul Leonard Edwardshow all
Deliberately lit fires make up over 15% of all fires in New Zealand buildings yet they are typically omitted from the design brief for fire engineering purposes. This report examines where deliberately lit fires should be included as part of the fire engineering design by examination of all deliberately lit fires recorded in the New Zealand Fire Incident Reporting System (NZ FIRS) between the years 1996 and 2006. The main types of buildings identified where consideration of deliberately lit fires within the design would provide benefits are: · Prisons · Psychiatric institutions · Schools · Crowd activities · Attached accommodation The report also examined what is required to include deliberately lit fires as part of the design process. Based on an analysis of the fire incident statistics, the majority of deliberately lit fires are the result of unplanned activities and existing design fires will be adequate. Two critical fire scenarios were identified as exceeding these requirements, the ignition of multiple fires and the use of accelerants. Greater life safety benefits are obtained by considering accelerants. In the case of multiple fires, each fire is likely to be within the capabilities of a fire engineered building however a number of such fires may overwhelm the fire protection features of a building. A number of issues for the fire engineer to consider are briefly discussed. In the case of accelerants, a number of experiments were completed to characterise the heat release rate and species production of a Molotov cocktail based on the fuel volume used. A second round of experiments extended this work by examining the scenario where a Molotov cocktail containing 1000 milliliters of petrol was deployed within a stairwell.