Elevated Fires

Sammanfattning: The primary objective of the present study is to evaluate the effect of the fire source elevation on the main fire parameters, by implementing laboratory experiments. In modern Fire Safety Engineering practice, it is often assumed that fire always commences at the floor level. However, it is not always the case. For instance, many residential fires were initiated by the ignition of ovens and stoves leading to the fire base to be raised above the floor by at least a meter. A series of tests with fire source located at different heights were conducted in the 1/3 scale of the ISO standard fire room. Heat Release Rate, Mass Loss Rate, temperatures inside the enclosure and radiative heat flux at the floor were measured. Results showed that when fires are elevated so that the flame is completely inside the smoke layer, they yield higher heat release rate, mass loss rate and heat flux than fires near the floor. At the same time, if the fire source is elevated to a level, where no considerable interaction with the smoke layer takes place, it will result in a less hazardous fire, as conditions critical to life safety develops slower than for the floor fire. The secondary objective is to establish whether the existing modelling software and analytical models are sufficient for simulating elevated fires. This is studied by comparison of the experimental results with simulation performed in FDS and hand calculations. Three simulations reproducing the experimental tests were performed using the FDS tool supplemented with an environmental feedback model. Simultaneously, hand calculations of HRR, MLR, HGL temperature and height were conducted by applying fire dynamics principles. FDS outputs and theoretical results were compared with the experimental ones. From the comparison it was learnt that current hand calculations and FDS models cannot sufficiently predict the impact of the source elevation on the development of fire parameters. The study provides speculations on factors that can be responsible for changes caused by the fire elevation. It also reasons why analytical and numerical models cannot account for these changes. However, no practical solution is provided in this work.