CONFERENCE PROCEEDINGS Australian Bushfire Conference, Albury, July 1999

copyright 1999

Abstract

1) Introduction

The southern states of Australia have long been troubled by bushfires. Seventy nine people died in the Black Friday fires of 1939, 62 in Hobart in 1967, and 72 on Ash Wednesday in 1983. Early firefighting efforts largely relied on manpower, with teams of men, often untrained civilian volunteers, using wet sacks, knapsacks and branches to beat out the flames. Present day efforts are largely mechanised, with 4 wheel drive tankers, bulldozers and aerial firebombers the tools of choice, but bushfire fighting remains a dangerous occupation. At least 52 Australian bushfirefighters have died on active duty since 1980, 10 in South Australia (CFS 1999), 19 in Victoria (CFA 1999) and 23 in New South Wales (RFS 1999). Excluding natural deaths, the majority of deaths have occurred when fire tankers have been overun by fire, and this continues to happen. Five firefighters died at Waterfall in 1980, 16 on Ash Wednesday in 1983, 3 at Grays Point in 1983, 1 at Wingello in 1998 and 5 at Linton in 1998.

2) Sources of Data

a) Actual Burnover cases

A search of the scientific literature and popular press yielded numerous reports of bushfire-fighter deaths extending back as far as 1939. The older reports, however, commonly related to "wet sack" firefighters and rarely distinguished untrained civilians from regular volunteers. This review is confined to burnover incidents involving regular organised firefighters occurring after 1979. Thirty-seven incidents were identified, 19 in Australia and 18 overseas. The Australian burnovers occurred as early in the season as September, and as late as April, and whilst a number occurred at major fires, others occurred during quiet times at small fires or burnoffs. Whilst the number of burnovers identified may seem disturbingly high, even these may represent only the tip of the iceberg, Almost every brigade contacted could recount a number of other "near misses". The Bridgewater Brigade, for example, had their tanker destroyed by fire on Ash Wednesday I in 1980, and the replacement seriously scorched on Ash Wednesday II in 1983, and again in 1990.

b) The Australian tests of 1972

In the early 1970s, a series of experiments subjected standard passenger cars to external heat simulating the effects of a bushfire (Cheney 1972). Even in the most severe tests, simulating a forest fire, the cars proved remarkably resistant to heat , it taking around 4 minutes for ignition of the door seals and roof lining, 8 minutes for the ignition of the engine bay, and over 30 minutes before the fuel tanks ignited. interior temperatures remained relatively low, reaching 104 degrees F after 4 minutes, but rising rapidly once the interiors caught fire.

c) The 1996 US Burnover tests

In 1996 the Missoula Technology and Development Centre conducted 3 separate trials which recorded the conditions inside fire appliances and personal fire shelters during an experimental burnover (Mangan 1997).

1. The first, the Florida Burn, involved a standard ‘Wildland Engine’ and a standard pickup with a slipon unit. The resulting low intensity grassfire, with low flame heights and an estimated intensity of less than 1500kW/m, passed over in less than a minute and did not seriously test the survivablilty of either the trucks or the fire shelters.
2. The First Los Angeles Burn subjected a standard urban appliance and a pickup to a more severe grass and brush fire with flame heights averaging 10-20 feet and an estimated intensity of 2000-3000 kW/m (Figure 1). The peak intensity of the fire passed within a minute. The urban appliance suffered severe damage in the fire, with cracked windows in the cabin, and ignition of paintwork, hoses, and mudflaps. The pickup was similarly affected, with scorched paint and cracked windows, and, disturbingly, filled with heavy smoke even though the interior did not ignite. The interior temperatures however, remained low, ranging from 60 degrees C in the pickup to 85 in the urban appliance, and were significantly lower on the floor of the cabins. The temperatures inside the fire shelter adjacent the urban appliance were higher than those inside the engine, reaching 220 degrees C at 12 inches above ground, and 160 degrees C at 1 inch above ground.

Conclusion Temperatures inside the vehicles remained survivable in this moderately high intensity burnover. Dense smoke however, possibly from near ignition of the door trim, may have forced the crew from the pickups’ otherwise survivable interior during the burnover. Temperatures inside the fire shelters were higher than in the vehicles.

3. The Second Los Angeles Burn was of lesser intensity than the first, with an estimated intensity of 2000kW/m , and with the peak of the fire passing within 30 seconds. The urban appliance again appeared to provide a survivable environment, with maximum interior temperatures of 75 degrees celsius (and only 40 degrees C on the floor). The pickup interior, however, became uninhabitable within 60 seconds when the interior trim of the door adjacent the fire ignited due to the radiant heat load on the exterior (Figure 2). This rapidly filled the cabin with smoke, and elevated the interior temperature to 280 degrees C. The temperature in the fire shelter adjacent the urban engine was below 80 degrees C. The shelter adjacent the pickup recorded a maximum interior temperature of 180 degrees C.

Conclusion Interior temperatures inside the fire shelters and the urban appliance remained survivable, (and again lower inside the vehicle) in this moderate intensity fire. Ignition of the interior door trim due to external radiant heat would however have forced the crew from the cabin of the pickup during the burnover.

4. The final burn, the Montana Burn subjected a Wildland engine and a pickup to radiant heat from an adjacent pine bonfire. This produced a more prolonged and severe exposure to heat than the earlier tests. Thick smoke and high temperatures made the Wildland Engine cabin uninhabitable after several minutes, again following ignition of the interior from the radiant heat. Conditions in the adjacent personal fire shelter however, remained survivable, with maximum temperatures of only 75 degrees C recorded after 12 minutes. The pickup, similarly, became uninhabitable after several minutes, when ignition of the interior caused the cab to fill with smoke and heat . The fire shelter adjacent the pickup recorded survivable interior temperatures of 105 degrees C.

3) Discussion of case reports and burnover experiments

a) Vehicle cabins : Windows and doortrims

Burnover reports from the US and Australia suggest that the large glass areas of modern vehicles are a major threat to crews sheltering in truck cabins. There are repeated references to severe radiant heat loads transmitted through vehicle windows, with crews resorting to holding up clothing or fire shelters to windows for additional protection. The Australian studies (Cheney 1972) showed that 50% of the radiant heat incident on a closed car window was transmitted through to the interior. In a least 2 cases, US crews were forced to abandon their vehicle cabins due to failure of the windows during a burnover. In addition, the Australian tests showed that ignition of the roof liners and door seals, which took around 4 minutes, even in their most intense "forest fire" burn, resulted in rising interior temperatures and dense smoke filling the cabin. Prior to ignition of the interior, the floor of the car remained a survivable environment. The US tests (Mangan 1997) showed that radiant heat loads on the outside of a vehicle door could cause its interior trim to ignite. This happened on 2 occasions after only 2 minutes in less intense fires. Following ignition, rapidly rising temperatures and toxic fumes again made the vehicle interiors uninhabitable. The earlier ignition of the door trims in these more modern vehicles may reflect the increasing use of plastics in modern vehicles. The survivors of the Carey Gully tanker lost in 1983 and the State Forests Tanker burned over in 1997 both described dense acrid smoke filling their cabins during the burnover.

b) Petrol tanks

The enquiry (Mc Arthur 1966) into the loss of 8 men at the Wandillo fire in 1958 reported that 2 of the 3 trucks involved experienced petrol tank ‘explosions’, and that the tank which did not explode was protected by a bash plate that also shielded it from radiant heat. The Australian tests however found that actual explosions were extremely hard to induce in vented petrol tanks because the vapour inside was too dense to ignite. At worst, heating of the fuel in the tank caused it to vent vapour which burned at the vent, but after the fire passed, this went out leaving unburned fuel in the tanks. Actual ignition of the tanks took over 30 minutes, even in severe test fires, and genuine explosions could only be induced in sealed containers, and then only after prolonged heating by fire. The fuel filler neck of Headquarters 81 reportedly burned off at Waterfall, and flames were observed around the filler cap of the Mylor Tanker as it was abandoned in 1980. It was subsequently recovered scorched but intact. The fuel tank of the Carey Gully Tanker still contained liquid petrol after the truck was burned out on Ash Wednesday II in 1983 (Figure 3). Sealed portable fuel containers, such as those used to store chainsaw fuels, may pose a greater risk than vehicle fuel tanks, particularly if stored in an exposed position . Ignition of the chainsaw fuel cans during the Grays Point incident reportedly contributed to the deaths of 3 crew sheltering near the truck. A similar ‘Jerry Can’ also exploded on the State Forests Tanker burned over at Gwabegar (Figure 4). The US tests were conducted with empty fuel tanks for the safety of the investigators. The greater reliability of diesel engines under severe heat and smoke conditions, rather than the lesser volatility of the fuel during a burnover, has now resulted in their adoption as standard for most bushfire vehicles (Carter 1994).

c) Flammability of other vehicle structures and equipment

Ignition of vehicle tyres, hosereels, fuel lines, air lines and other external equipment may contribute to the entrapment of the vehicle, or its subsequent destruction by fire. During the Second Los Angeles burn, the mudflaps and hose reels on the urban appliance ignited, and the airlines to the brakes burned through, immobilising the vehicle. In another real world incident, a New South Wales Tanker was nearly lost when a small spotfire which ignited under the vehicle burned through the air line, locking the brakes and preventing it being driven clear. The crew of the Glen Park Tanker found the rear tyres of their appliance alight after the Creswick burnover, and were unable to access their knapsack sprays due to their storage position, externally, adjacent the burning tyres. Fire subsequently destroyed the truck.

d) Mechanical failures

Failure of truck engines to start, or continue to run, under severe heat and smoke conditions are common factors in burnover incidents. The petrol engine of the Mylor Tanker stalled after it was driven through flames at Longwood in 1980, forcing the crew to abandon it on the roadway, although without injury. It was seriously scorched as the fire passed over, but driven away the next day after cooling down. On Ash Wednesday 2, the Tea Tree Gully Tanker was abandoned to the approaching fire after the rear differential failed, and the Cudlee Creek Tanker was abandoned after its petrol engine failed, apparently due to fuel vaporisation. Both vehicles were destroyed by fire and one crew member was injured. In 1985, Santa Barbara County Engine 13 was engaged in protecting a house from an approaching urban interface fire. Conditions worsened, and the decision was made to retreat , but intense smoke and embers choked off its diesel motor and burned out the turbo. The crew were evacuated safely in another vehicle. Linda Engine 3 was destroyed at the 49er fire in 1988 when an electrical fault prevented the PTO pump from being disengaged. The crew escaped on foot. The crews of multiple fire engines sheltered in their cabs during a prolonged (but not very intense) burnover in a safety zone in a forest at the Marre Fire in 1993, with Engine 2-7 suffering engine failure when its diesel engine ingested burning embers. In 1995, the petrol engined Kuna Engine 620 was already disabled on the fireground, probably with an engine overheat, when it was over run by an intense, wind driven brush fire and the 2 crew, sheltering in the cabin, were killed. Engine 97 was lost at the Fayette fire in 1996 when its engine stalled and would not restart as the driver tried to reverse out of flames when a backburn flared up with a wind change. The diesel engined Los Angeles City Engine 10 stalled after being stopped in traffic in heavy smoke at the 1996 Calabassas Incident. The truck was then burned over whilst the crew of 4 sheltered under part opened personal fire shelters in the cabin. One (not wearing PPE) received serious burns, the others minor. The Glen Park Tanker was destroyed at the Creswick fire in 1997 when overrun by the parent fire whilst fighting a spot fire when its diesel engine stopped (apparently due to an incidental electrical fault, rather than ingestion of heat and smoke) and would not restart.

e) Circumstances leading up to entrapment

Table One summarises the principal cause of the entrapment. A sudden and unexpected intensification of the fire, usually due to a wind change, is reported in nearly every case of burnover. Often there is very little warning before the crew is overrun, leaving no time to escape. In some cases the burnover occurs so suddenly there is no time even to position the vehicle, and the burnover must be ridden out in whatever location and orientation the vehicle is in. In 1980, near Waterfall, Headquarters 81 was trapped by a finger of fire coming out of a gully in a "sudden blowup", with the loss of all 5 crew. The survivors of the Carey Gully Tanker, lost on Ash Wednesday in 1983, reported "3 seconds warning before a fireball hit". The crew of the Echunga Tanker reported that "20 foot high flames came out of nowhere" when they were also caught elsewhere on the same day. Linda Engine 3 was protecting a house from an approaching brushfire when "a sudden increase in the wind" resulted in 125 ft high flames. At the 1989 Eagle fire, Engines 2387,4474 and 2379, with most of their crew dismounted to battle a spotfire, were suddenly overun by a fire whirl generated by the mainfire. This whirlwind crossed the intervening freeway and with such speed and ferocity that one firefighter was unable to climb into his vehicle despite the fact that he was already holding the doorhandle when the incident occurred. Several other crew were bowled over and burned, and only one was able to deploy a fire shelter. Kuna Engine 620 was already disabled on the fireground at the 1995 Point fire, when a windshift caused a sudden intensification of the fire, overrunning the truck and killing its 2 crew. Engine 97 was patrolling a backburn at the 1996 Fayette fire when "the wind increased’’ causing flames to impinge directly on the truck. The engine stalled, despite repeated attempts to restart, forcing the crew to bailout and leave the truck to be consumed by fire. Intensification of the fire, which travelled "over 600 feet in 30 seconds" as it ran uphill out of a bowl, contributed to burnover of the 3 appliances in the California’s "Calabassas Incident" in 1996. Three Bulldozers and a utility constructing a break were destroyed at the Logan fire in 1997 when the fire "made a rapid run at them, cutting off their escape route", the crews were lifted out by a helicopter which was fortuitously overhead at the time. Engine 3473, patrolling a back burn at the same fire, survived being overun by fire when changing conditions produced spotting. The deaths of 2 highly experienced firefighters, mopping up on Scotsman’s Hill in 1997 followed a hot windshift in which the "fire moved hundreds of metres in seconds". The loss of the Wingello Tanker, with one death and seven injuries, followed "an uncharacteristic wind change of uncharacteristic strength" at a backburn. The survivors of the 1998 Linton fire, in which the Geelong West Tanker was destroyed with the loss of all 5 crew, reported a sudden flare up on a quiet fire following a windshift, "within about 2 seconds, we were fully engulfed with flames coming right over the cabin, the wind was horrific".

f) Self defense spray systems

Some of the vehicles involved in burnover incidents had functioning hand held or fixed self defense spray systems in operation. The survival of the Geelong City Tanker, and the loss of the Geelong West Tanker, at Linton in 1998 graphically indicated the potential benefit of such systems. In this incident, 2 essentially similar tankers were overrun by fire on a trail in heavy bush following a wind change. The crew of the Geelong City Tanker took shelter in their truck, operating fog lines for self defense. Their available 1000 litres of water was exhausted just after the fire passed, and the truck and crew survived with superficial damage. The Geelong West Tanker, however, was out of water and was destroyed in the fire, and all 5 crew were killed (Figure 5). Earlier in the day, the Snake Gully Tanker crew exhausted their available 700 litres during their appliance burnover at the same fire, though the vehicle survived with severe scorching. Engine 3473, patrolling a back burn at the 1997 Logan fire, survived being overun by fire when changingconditions produced spotting in a forest. Two crew defended the truck with a fogline from the wheel arch as the fire burned past. The truck was externally scorched but there were no injuries. When the Glen Park Tanker was overrun, 2 crew sheltered under a blanket in the cabin whilst the other 3 sheltered in the rear crew haven, operating hoselines for protection whilst the fire passed over. All survived uninjured, though the remaining water supply of approximately 1000 litres was exhausted in the burnover and the rear tyres caught fire, subsequently destroying the truck. When Glendale Engine 24 was overrun by fire in California, one crew member remained with the vehicle, using a hose to knock the fire down in the trees as it went past. He survived without injury. Three others ran for their designated safety zone in a nearby house. One was critically burned and the other 2 sustained minor burns as they ran. The Bridgewater Tanker destroyed at Longwood on Ash Wednesday 1980 was reversing down a road away from the fire crossing in front. It ran off onto the road shoulder, hit a small tree and stopped. A branch from the tree struck the pump throttle, disabling the on board sprinkler system. The crew of 5 all sheltered within the truck cabin during the burnover, but all 5 sustained moderate burns contacting hot metal bailing out after the fire. The crew of the Isuzu Tanker burned over at Ridgeway in 1990 pumped their entire 3600 litre water supply through their self defense sprinklers as the fire passed, then bailed out to shelter behind a house for the remainder of the burnover.

4) What Conditions are Survivable During a Burnover?

The principle threat to survival during a bushfire is radiant heat. Actual air temperatures are often quite low, particularly near the ground (Cheney 1972). The maximum conditions survivable depends on the duration of the exposure and the protection available. One model suggests a maximum air temperature of 100 degrees C and a maximum body temp rise of 2 degrees C as survivable (Knight 1988). Another (Bond 1986) suggested a maximum 15 minute exposure of 120 deg C. Higher air temperatures may be survivable for shorter periods during a vehicle burnover provided there is adequate radiant heat protection.

A typical low to moderate intensity grass fire may produce flame heights of 2-3m, intensities of around 1500 kW/m and pass by in less than a minute. The Florida Burn (Mangan 1997) suggested that such a fire is easily survivable in either a fire shelter or a vehicle. There are also reports of survival in such fires by firefighters who laid on the ground or ran back through the flames onto the black. Fatalities have however occurred amongst crews who abandoned their vehicles or were caught in the open. A more intense grass or brush fire, producing 6m flames, and intensities of 3000-4000kW/m, similar to the Los Angeles Burns, (Mangan 1997) passes by in less than 2 minutes and may be survivable in shelters or vehicles. In one of these burns, conditions remained survivable in one vehicle, though ignition of the door liner might have forced a crew to evacuate the other during the burnover. Cheney has suggested that 2000-3000kW/m is the maximum intensity fire a tanker crew will survive in their truck during a burnover. This may however be an underestimate, for some crew survived in each of the Carey Gully, Linton and Wingello burnovers, all of which occurred on narrow firetrails in heavy bush and may have been more severe than this (Figures 3,5 & 6). A more intense fire, such as a forest or bush fire may involve flame heights of more than 10m, intensities of more than 5000-10,000kW/m and take 3-6 minutes to pass by (Bond 1986). The surrounds may then remain hostile for another half hour as the larger fuels burn out. Survival may still be possible with suitable protection in an appropriate clearing. Certainly there are numerous reports of survival in shelters in clearings during US forest fires. Most of the Australian appliances destroyed by burnover have been lost on narrow trails in bushland and some crew have survived. The Australian tests indicated that conventional passenger cars provided survivable conditions for approximately 4 minutes in a severe fire, equivalent to a forest fire. Following experiments carried out in Australia in 1988, Knight (Knight 1988) estimated that the maximum fire intensity survivable in a personal shelter in a 10m clearing was 5600kW /m.

5) Improving the Protection Provided by Bushfire Appliances

a) What are the current protective systems?

Current appliance protective systems include high sides, crew havens, fixed sprinklers, hand held hoselines and radiant heat curtains. The present generation 2000 and 3000 litre bushfire appliances used by the South Australian Country Fire Service (CFS) are built on standard commercial 4WD truck chassis’. The crew of 6 is split equally between the drivers cabin and a ‘crew haven’ on the tray immediately to its rear (Figure 7). In the event of a burnover it is intended that 3 will shelter in the cabin and 3 in the crew haven. The drivers cabin is a standard commercial truck cabin and has no additional heat protection. Low volume sprays aremounted on the truck bumper to cool the front tyres but are not intended for cabin protection. A woolen blanket is carried in the cabin for crew protection. The rear crew haven is a purpose built structure intended to provide both radiant heat and rollover protection for the 3 crew on the rear. It has a heavy steel frame, bench seating for 3 with lap belts and a roll down reflective heat shield. It is clad in fibreglass and has equipment lockers outside, giving double layer protection for the sides. Two short 25mm self defense hoselines, plumbed to the main water supply of the truck and used for fire suppression on the move, are available at this position. In the event of a burnover it is intended that the 3 rear crew will shelter inside the crew haven with the reflective screen down and operate the 2 self defense hoselines for additional protection. A similar system, minus the roll down screen, was used on the Geelong City Tanker which survived the Linton Fire.

b) The vulnerabilities of present tankers

The US tests , and the analysis of the burnover reports, suggested that whilst the cabin of the truck can provide good protection for the crew during the burnover, the windows and doors are vulnerable to the effects of radiant heat. There are numerous reports of personal fire shelters being held up to windows to block the influx of radiant heat. The provision of a drop down or pull around radiant heat screen inside the vehicle cabin would be an inexpensive and worthwhile improvement. The South Australian Country Fire Service is presently considering such a screen (Figure 8). The door skins themselves appear to be unexpectedly vulnerable. The effects of radiant heat on the outside of the door produced early ignition of the door liner in 2 of the US tests. In each case, this resulted in the cabin rapidly filling with heat and choking smoke which would have forced the premature evacuation of the cabin during the burnover. This phenomenon may become increasingly important with the increasing use of plastics in modern vehicles. Improving the radiant heat protection for the vehicle doors, possibly by applying a reflective coating, or a clip on screen, akin to a ‘tropical roof’, is a major priority. The crew haven presently fitted on the working tray of South Australian Country Fire Service appliances probably provides excellent radiant heat protection, indeed, given its lack of windows and door trims, it may provide significantly better protection than the driver’s cabin. There is a trend in Australia towards the adoption of dual cab designs for Bushfire appliances, with the elimination of the crew haven. This may be a retrograde step in terms of radiant heat protection, particularly if the vulnerabilities of the windows and door trims are not addressed. It should also be noted that many burnovers occur with little warning, and adequate crew protection must be provided for both the cabin and the working area on the back of the appliance. The use of handlines or dedicated self defense sprinkler systems to project a curtain of water as a radiant heat shield between the vehicle and the fire clearly has significant potential to improve survivability in the more severe burnover incidents. Hand directed hoselines may be more effectively directed towards the approaching fire, but fixedsprays allow the crew to take cover. In limited trials here, the presently recommended strategy of retreating to the safety of the crew haven and blindly directing a hand held hose from within, proved ineffective, as the spray tended to favour the rear of the truck and provide limited protection for the crew haven and none for the drivers cabin. Both systems are reliant on a supply of water and the continued operation of the pump. Tanker designs presently in use in South Australia provide inadequate protection for the pump. A variety of fixed sprinkler systems are undergoing wind tunnel evaluation in New South Wales at present.

Analysis of the US tests, and the cases above, would suggest that such systems should :

• be capable of operating for at least 2 minutes on the dedicated water reserve
• be easily operated from both the drivers cabin and the working area on the tray
• be capable of operating effectively in high winds
• operate effectively at all ranges of pump delivery from idle to full power.

In limited trials with a variety of rotary head and venturi sprinklers we noted :

• it is difficult to provided adequate cover with a single, vertically directed nozzle as such a set up is excessively affected by winds
• the use of multiple smaller spray heads inclined more to the horizontal than the vertical markedly improved the cover provided by the spray.
• rotary head sprinklers were more affected by variations in supply pressure, producing an excessively fine spray at high supply pressures. In addition they may be more prone to failure due to failure or jamming of the moving parts

As long ago as 1958, the coroner noted the vulnerability of the petrol tanks of bushfire appliances to fire. The lesser volatility and greater reliability of diesel fuel has lead to its increasing adoption as the fuel of choice, and there are no identifiable reports of diesel tank ignition in survivable burnovers. A degree of heat shielding for these tanks may however be prudent.

6) Burnover Incidents in Personal Fire Shelters

Personal fire shelters, in the form of lightweight reflective ‘A tents’, were first developed in the 1950’s (Figure 1). Early failures in testing led to their withdrawal from the Australian market, although

they have since been widely adopted in the United States. Personal fire shelters have been mandatory equipment for Wildland firefighters in the US since 1977 and have been credited with saving nearly 300 lives. When threatened, the firefighter pitches the shelter on a patch of barren ground and lies face down inside, holding the edges down with the handholds. With practice, a shelter can be deployed in around 30 seconds. Fire shelters have been deployed many times in the US, most often when ground crews are caught by fire in forests. They appear to provide good protection in moderate intensity burnovers, but may not ensure survival in a severe fire. Fatalities also occur if fire overtakes the crews rapidly, when conditions are too windy for deployment, or a suitable clearing is not available. It should be noted that whilst personal fire shelters are excellent radiant heat reflectors, most are not flame proof and require an area of clear mineral earth for deployment. Such safety zones must be prepared in advance, or naturally occuring, for there is unlikely to be time to prepare them once threatened. This makes them an unreliable protective device in many fast moving Australian grass and bushfires.

7) Does a Personal Fire Shelter Provide Better Protection Than a Fire Appliance?

There is no clear answer as to whether, given the choice, a firefighter should ride out a burnover in the appliance, or in a personal fire shelter. Temperatures tend to reach higher levels, though for shorter times, inside fire shelters than in vehicles. The vehicle may provide better protection, provided premature ignition of the door liners does not occur. It seems, however, that fire crews often appear to have little choice. Bushfire fighting in Australia is largely mechanised, and most burnovers have involved firetrucks over run with little warning, forcing crews to ride out the fire wherever they are. There are however instances in which circumstances have prevented crews from sheltering in their vehicles, or forced them to abandon them during a burnover, and bail out into a still hostile environment. When the Carey Gully Tanker was overrun at Mount Bonython, 3 of the crew were unable to return to the vehicle as it was already immersed in flames. They were forced to flee on foot. One successfully outran the flames, but the other 2 were forced to seek shelter whilst the flame front passed. The Captain described crouching in a clearing and turning like a chicken on a rotisserie to stop the side facing the fire from getting too hot. He was seriously burned. The other fireman was burned to death in a culvert nearby. Two men died at Lithgow in 1997 when a sudden flare up caught them away from their vehicle. The survivors of the State Forests Tanker used blankets to shield themselves from the heat after heat and smoke forced them to bale out of their tanker after a burnover. The crew of the Wingello Tanker sheltered in the crew haven of their appliance until forced out by intense heat. They then described kneeling on the ground in an environment that was still so hot it caused further burns. When Engine 2387 was suddenly over run by a fire whirlwind in California, 5 crew were caught away from their vehicles. One successfully deployed his shelter and survived without injury. Four others were burned, one critically. Even in the US, where personal fire shelters are mandatory for all wildland firefighters, crews continue to be caught unexpectedly and forced to take shelter in their trucks. In many instances, personal fire shelters have been used as additional radiant heat shields by crews taking shelter in vehicles. Conversely, most shelter deployments in the US have involved ground crews with no other form of protection. Ground based firefighters have no other options, they should receive fire shelters and appropriate training. Mobile crews should have appliances with optimal inbuilt burnover protection. In some circumstances, appliance crews may derive additional protection if they also carry personal fire shelters.

8) Conclusions

Being overtaken by fire remains a significant cause of death for Australian bushfire fighters. Often the burnover occurs so suddenly there is little time to prepare, and it must be ridden out in whatever location and orientation the appliance is in. The protection afforded by present appliances is not optimal, and can be significantly improved by the fitting of :

• radiant heat shields to the exterior of the drivers’ cabin
• heat reflective curtains inside the cabin
• properly engineered fixed self defense sprinklers, with appropriate water reserves
• radiant heat protection for the pump

In addition, reasonable steps should be taken to minimise the flammability of exterior structures and to reduce the vulnerability of vital systems, eg brake lines. Australian bushfire fighting is largely mechanised and most burnover incidents have involved mounted crews. Bushfire appliances should be engineered for maximum crew protection. Personal fire shelters may still prove useful if crews are caught away from their vehicles, require additional protection during a severe burnover, or are forced to bail out into still hostile environments during or after a burnover.

9) References

Bond A and Cheney NP. (1986).

A discussion paper on techniques and equipment for bush firefighters entrapped by fire. (CSIRO National Bushfire Research Unit), Dec.

Carter E and Milton B. (1994).

Internal combustion engine performance on the fireground. Int J Wildland Fire 4(2):83-91

CFA. (1999).

Country Fire Authority of Victoria. 8 Lakeside Dr, Burwood East, Victoria.

CFS. (1999).

South Australian Country Fire Service. 20 Richmond Rd, Keswick, 5035.

Cheney NP. (1972).

Forestry and timber bureau studies human behaviour in bushfires… don’t panic and live. Nat/Dev Sept (CSIRO Division of Forestry.)

Cheney NP.

The safety of Bushfire fighters.

Knight I. (1988).

What intensity of fire can a firefighter survive in a reflective shelter. Fire Technology Nov. 312-331.

Mangan R. (1997).

Surviving Fire Entrapments. Report no 7E62P87, US Department of Agriculture Forest Service. ( Missoula Technology and Development Centre, Montana.)

McArthur AG, Douglas DR and Mitchell LR. (1966).

The Wandillo Fire, 5 April 1958. Forest Research Institute. Forestry and Timber Bureau. (Commonwealth of Australia, Canberra).

RFS. (1999).

New South Wales Rural Fire Service. Unit 3, 175-9 James Ruse Dr, Rosehill, 2142.

10) Acknowledgements

Thankyou to the many people who helped with this paper, especially the burnover survivors who consented to talk about their experiences, and to Arthur Tindall, Helen Pereira, Martyn Kiellor, Jason Greenlee and Dick Mangan.