A theoretical model of the area required for safety zones where a fire fighter dressed in Nomex protective clothing could survive the radiant heat load has been prepared by Butler and Cohen (1998). This model suggests that safety zones should be 4 times the expected flame heights eg around 40 m for 10 m flames in grassy fuels. Forest fires, where flame heights may exceed 50 m, require safety zones in excess of 200 x 200 m. These calculations did not include the effect of convective heat transfer which can be significant, particularly on sloping ground.
However they should provide satisfactory guidelines where people have some protection from convected heat, e.g. inside a vehicle.
Whether or not protective shelters are provided for hand crews or tanker crews authorities must seriously consider the area needed for effective deployment, the time required to clear it, and the training needed to prepare firefighters for what they are likely to endure.
Area: Fire shelters must be deployed on an area of mineral soil cleared of all combustible material. The size of area depends on the type of fuel and the intensity of the fire. Knight (1988) recommended a 10 x 10 m fuel-free clearing for eucalypt fires up to 5000-6000 kW m-1 (crown fire in 30 m mixed stringybark forests). Butler and Cohen (1998) calculated that protective shelters would provide protection in areas where there was 15 m separation between the shelter and the flames. The model did not include the effects of convected heat and is under revision (Butler 1998).
Under extreme conditions, or on steep slopes, or in heavy fuels, I am confident that areas have to be much greater in order to reduce the risk of violent blasts of hot air ripping the tents away from the occupants and to allow dilution of hot combustion gasses by entrainment of surrounding air.
Rates of Clearing: The average rate of clearing fire line in eucalypt litter and low shrub fuels by a trained, fit firefighter is around 2 m2 per minute (Budd et al. in preparation). This rate is similar to measured rates for clearing 2 x 1 m2 shelter sites in hardwood litter in North America (Bennett 1990). Therefore, it would take a firefighter 10 minutes to clear a 5 x 4 m site similar to that used by King or something approaching an hour for a 10 x 10 m site recommended by Knight. Clearing sites with substantial grass or scrub material on them would take considerably longer. Clearly it is not possible to clear a substantial area suitable for surviving high-intensity fires in a few minutes.
Training: In the US fire shelter training is mandatory for all firefighters. Putnam (1991) recommends a 4 hour course which includes coverage of Standard Fire Orders and Watch Out Situations, case studies, visualisation of entrapments and practice deployment (without fire). Deployment on test fires has resulted in firefighters receiving burns and narrowly avoiding serious injury and death and is strongly discouraged by the USDA Forest Service (Mangan, 1993).
In practice the recommended training for new firefighters is not always carried out (Ian Dicker pers. comm.) and annual refresher courses are missed completely (Rosenkrance et al. 1994). I have serious reservations about issuing a piece of equipment then not providing realistic "hot" training sessions. Like many aspects of fire training, firefighters must experience for themselves the performance of the equipment under moderately severe fire conditions before they can properly assess the relative importance of actions to be taken during wildfires. Book or classroom exercises may not always be clearly interpreted; for example one American text on wildland fire illustrates a fire shelter pitched on continuous fuels (see Pyne 1984).