A stochastic model for assessing bush fire attack on the buildings in bush fire prone areas

From WikiPapers
Jump to: navigation, search

A stochastic model for assessing bush fire attack on the buildings in bush fire prone areas is a 2009 conference paper written in English by Tan Z., Midgley S. and published in 18th World IMACS Congress and MODSIM09 International Congress on Modelling and Simulation: Interfacing Modelling and Simulation with Mathematical and Computational Sciences, Proceedings.

[edit] Abstract

Bush fires are a major natural and socio-economic hazard in Australia. Under extreme fire weather conditions, bush fires spread very rapidly and are difficult to contain by firefighting services. When spreading on the rural/urban interface, they can cause significant damage to buildings or structures. The well known examples of such disastrous bush fire events include the bush fires which occurred in Tasmania in 1967, Victoria and South Australia in 1983, New South Wales in 1994, Canberra in 2003 and Victoria in 2009. The number of houses lost as a result of these fires is over 1300, 1500, 200, 500 and 2000 respectively (Leonard & MacArthur, 1999; Ellis et al., 2003; Blanchi & Leonard, 2005; Wikipedia 2009). To minimize the risk of building loss from such devastating bushfires, many bushfire protection measures have been developed and implemented within each State in Australia. One of the most effective and commonly used measures is the application of construction and design standards to developments in bushfire prone areas. However, the appropriate application of this protection measure requires the use of a bushfire attack assessment model to determine the level of bushfire attack to which a development might be exposed based on the site specific variables associated with weather, fuel and topography. At present, almost all the existing bushfire attack assessment models available for use are the so-called deterministic models (Ellis, 2000; Tan et al., 2005; SA, 2009), which are based on radiant heat flux modelling. The principles of these models are the same, i.e. taking deterministic values for all the input variables and producing the deterministic output of radiant heat flux. In situations where there exists a significant level of uncertainty with the inputs required by these models, it may be difficult to choose the appropriate values for them and therefore the risk level associated with the output on which a decision is made is usually unknown. This means that the safety levels of the decisions based on the deterministicmodels' outputs may be either more than adequate, due to the use of conservatively high values, or inadequate due to the use of the conservatively low values for the inputs with uncertainties. In view of the above, a stochastic bushfire attack assessment model has been proposed by the Authors. The principle of the proposed model is that the model's output i.e. radiant heat flux is calculated repetitively with the randomly sampled values for the inputs with uncertainties using Monte Carlo sampling. The model output is not a single radiant heat flux but a radiant heat flux probability distribution reflecting the uncertainties with the model inputs. Based on the radiant heat flux probability distribution, the radiant heat flux for a given percentile or safety level and the corresponding standard construction requirements can then be determined. Therefore a risk based decision in relation to the application of appropriate standard construction requirements to a development in bushfire prone areas could be made. The implementation of the proposed model makes use of a commercial software product called @Risk, which involves a number of steps like developing @Risk spreadsheet model, analyzing the model with Monte Carlo simulation, determining radiant heat flux for a given percentile or safety level and determining the level of bush fire attack and the associated standard construction requirements as per AS 3959 (SA, 2009). The use of the model has been demonstrated by an application example. As demonstrated in the example, the major advantage of the proposed model over the existing deterministic models is that the construction standard determined by this model for a given development could be based on a known minimum safety level. This approach provides construction standards for the proposed development which are likely to be more cost effective whilst providing for pre-defined safety levels.

[edit] References

This section requires expansion. Please, help!

Cited by

Probably, this publication is cited by others, but there are no articles available for them in WikiPapers.