SUMMARY. On 18 March 1996, in a tunnel along the Palermo-Punta Raisi motorway (Italy), a tank-truck carrying 2500 litres of liquid petroleurn gas (LPG) was involved in an accident. The truck was hit by a coach, which ripped off the top of the coach, provoking the escape of gas, which caught fire within seconds. The flames heated the LPG still in the tank, causing a BLEVE(boiling liquid expanding vapour explosion). At the moment of the accident there were 22 cars and a minibus in the tunnel, besides the coach, with a total of 50 persons. The delay between the ignition of the gas and the explosion enabled numerous passengers to escape. However, five people were burned to death and twenty were severely injured. Ten people presented minor burns or other traumas. The following aspects are analysed: the dynamic of the accident, the dynamic of the rescue operations, the management of the severely injured patients treated at the Palermo Burns Centre, the psychological reaction of the population, relatives and injured persons, the evolution of the pathologies, and the medical and surgical treatment. The Palermo tunnel disaster, with its terrible dynamic, confirms the thesis that the only way to counteract the impossibility of predicting this type of accident is through the coordinated response of rescue aid at all levels, from the general population to physicians, nurses, volunteer groups, Civil Defence, police, etc., who must all be adequately trained in appropriate education campaigns to prevent fire disasters. This accident, the first case described of a BLEVE in a closed and restricted environment, can provide material for indications and suggestions for the formulation of research hypotheses, operative proposals and teaching possibilities, with a view to realizing an ever higher level of disaster preparedness.

Introduction
On 18 March 1996, on the motorway between Palermo and Punta Raisi Airport, there was an accident in a tunnel in which a tank-truck carrying liquid petroleum gas (LPG) caught fire and exploded, leading to a fire disaster that caused the death of five persons and burn injuries in 34 other persons.

The scene of the disaster
The tunnel, 148 metres long, carries one-way traffic in the direction Punta Raisi-Palermo. At the midway point there is a bend to the right of about 25'. The road surface in the tunnel slopes from left to right to assist the drainage of rainwater. At a distance of about 100 metres from the entrance of the tunnel there are signs imposing an 80 kph speed limit. The tunnel is not provided with illuminated warning signals. Near the entrance to the tunnel there are signs advising drivers of the presence of radar speed-check devices.
Parallel to the tunnel is another one the same length carrying traffic in the opposite direction (Fig.1). The two tunnels are connected by a passageway measuring 5 x 10 m. The tunnel is 10 km from Palermo and 7 km from Punta Raisi Airport (Fig.2). The nearest motorway exits are Carini, Capaci, and Isola delle Femmine. The nearest towns with hospitals are Carini (4 km), with a small hospital, and Palermo (10 km), with a Burns Centre and the Villa Sofia and G. Cervello Emergency Hospitals.

Fig.1 - The two parallel morway tunnels involved in the fire disaster. Entry from Punta Raisi. The fire was in the right tunnel. Fig.2 - Map showing distances between Palermo, Punta Raisi, and nearest small towns

Every day approximately 40,000 vehicles use the motorway each way, with peak traffic in the summer period and at certain times of day (up to 10,000 vehicles per hour). At the time of the disaster (2.30 p.m.), traffic was fairly busy as many people were returning from the country into town, and vice versa. In particular, there were several school buses, and a number of charter flights had just landed at the airport.

It was raining on 18 March 1996. At about 2.20 p.m., either because the road was wet or because of excessive speed, a car skidded, hit the guard-rail, and turned over, blocking the lane. Other vehicles were following, including a minibus. There was a massive pileup and the whole tunnel was blocked (involving sixteen vehicles in all).
At the same moment a tank-truck approached, carrying a load of 2,500 litres of LPG. The driver succeeded in stopping the truck without crashing into the stationary vehicles. He immediately activated his flashing amber light to warn following traffic of the danger.
However, a Mediterranean Tours Coach with fourteen persons on board was not far behind. The driver was unable to stop the coach, which skidded to the left' and crashed violently into the tank-truck. Immediately behind the coach were four other vehicles, which crashed into the coach and into each other. At this moment nineteen cars, a minibus, the coach and the tank-truck were stuck in the tunnel, with some 50 persons on board (Fig. 3).

Fig. 3 - Reconstruction of the disaster in the tunnel

The impact between the coach and the tank-truck was violent. As a result there was a gas leak from the tanktruck due to damage to the upper part, which had been ripped open (eye-witnesses remember a plume of white vapour above the tank, and when the various parts of the tank were reassembled it was found that the upper part was ripped open).
After a few seconds, according to witnesses' reports, there was a first explosion. Probably the cloud of gas ignited as a result of a spark from a car motor that was still running or because of contact with a hot engine or with gas mixed with air. Witnesses remember being invested with a hot blast that caused a burning pain in exposed parts of the body, without causing much damage to the vehicles, where most of the people were.
The fire mainly involved the front part of the coach, which caught fire. The driver later reported that he saw a sudden large flash and black smoke outside the coach, at which point he shouted to the passengers to move to the back of the coach. He was unable to open the passenger doors. Someone broke open the rear window, through which passengers escaped. The tunnel was dark and full of smoke, and it was not noticed that five persons had been left in the coach, possibly overcome by the fumes.
The fire continued to feed the burning gas escaping from the tank. After 6-7 min, when everybody had escaped from the tunnel (except for the four persons left in the coach, later found completely burned, and one person who was discovered between the coach and the tank~truck), there was a tremendous explosion with violent flames and black smoke, and shock waves at the exits of the tunnel, and also of the adjacent tunnel (Fig. 4).

Fig. 4 - The violent fire in the tunnel after the BLEVE

The second explosion was presumably a case of BLEVE (boiling liquid expanding vapour explosion) caused by the direct action of heat on the tank. This would then have led to the boiling of the liquid gas, the formation of more gas, an increase in pressure, and further ripping open of the tank, with the possible leakage of LPG still in liquid state, which immediately evaporated." The cloud of gas and the drops of LPG caught fire immediately, causing the explosion of the tank (Figs. 5).
How does a BLEVE occur?

Subsequent examination of the four pieces of the tank found after the disaster confirmed that the explosion occurred inside the tank, in the central part (Figs 7a, b, c). The fire then spread all over the tunnel and was finally extinguished by the fire brigade after two hours (8a, b, c, d).

Fig. 7a - The tank-truck after explosion. Fig. 7b - Pieces of the tank-truck after explosion. Fig. 7c - Pieces of the tank-truck after explosion.

Fig. 8a - Aspect of tunnel after fire. Fig. 8b - Aspect of tunnel after fire. Fig. 8c - Aspect of tunnel after fire. Fig. 8d - Aspect of tunnel after fire.

The BLEVE lasted for 5-6 seconds, reaching a temperature of over 1000 °C. Witnesses remember that after the collision between the cars a fire started in one of the vehicles (the tank-truck driver said that he tried to extinguish it with a fire extinguisher) and that there was an explosion a few seconds after the crash between the coach and the tank-truck (first explosion). Six or seven minutes later, when the survivors had left the tunnel, there was a violent explosion, with great flames at the exits of the tunnel and a sharp shock wave and smoke, also in the other tunnel (BLEVE).

Situation at the Burns Centre at the moment of the disaster; emergency preparedness

The Palermo Burns Centre has twelve beds for intensive care (intensive area, IA), plus eighteen beds for patients with less serious burns, in the recovery stage, or having reconstructive surgery (post-intensive area, PIA), adjacent to an emergency Operating Room (Fig. 6). The Centre is on the second floor of a new building, of which the first floor is occupied by the Plastic Surgery section (50 beds). The intensive care beds are located, in pairs, in six rooms, each with its own autonomous climatization plant.

Each patient is monitored visually from a nurse-station by means of a TV circuit. Monitoring of heart function, pressure, respiration and weight is also centralized and controlled from the nurse-station. All data can be computerized.
The intensive care section is surrounded by an external corridor from which visitors can view the patients through a window and speak by means of an intercom system. This section is self-sufficient - it has its own laboratory for chemical analyses and its own pharmacy. It is in direct contact with the operating theatre complex by way of a stretcher lift.
The eighteen PIA beds occupy the other half of the floor.
The third floor is taken up by the operating theatre complex for the whole Department. There is also a Modumed operating room with a passe-malade reserved for the surgical treatment of burns, and there are two other operating rooms for the activity of the Plastic Surgery Sections.
On the ground floor is the Emergency section of the Department, with two operating rooms for emergency surgery, out~patient rooms, offices and the library. The rehabilitation rooms are in the basement.
With its up-to-date technology, the Palermo Burns Centre is generally recognized as one of the the most modern in Italy.
At 2.40 p.m. the Burns Centre was alerted to the possible arrival of a number of burn patients (Table I).

As soon as news of the disaster reached the Burns Centre (2.50 p.m.), the emergency plan for such an event immediately became operative. At that particular moment there were three physicians on duty present in the Department, plus another physician on call in the afternoon and evening. The Chief of the Department was immediately warned, and six other physicians were put on pre-alarm status. Eight IA beds were made available, as well as three PIA beds. The head nurses in the IA and PIA sections were warned and the nursing teams in the two sections were reinforced.
All the sterile linen in the Department was transferred to the Burns Centre, together with all other material that might be useful in handling the arrival of a massive numher of patients.
The hospital pharmacy was instructed to send quantities of Ringer's lactate, acetate and PPS. The central Emergency Department of the hospital was instructed to send a physician to carry out the necessary procedures for admission of the patients and to perform anti-tetanus prophylaxis.
Of the four physicians present, two remained in the Emergency rooms on the ground floor and two in the Burns Centre.
The first two patients arrived at 3.40 p.m. (50% and 15% burned BSA, respectively). Both had burns in the face, hands and other parts of the body. They were immediately taken to the Burns Centre Emergency section. Both said they had inhaled hot air and smoke. The patients were examined to check percentage BSA and the degree of their burns. Infusion treatment and medical therapy were initiated immediately. The patients were admitted to IA.
By 5.30 p.m. all sixteen patients had been monitored and subjected to lab tests and thorax radiography (Tables II, III)

Discussion

On the basis of an analysis of the factors that combined to cause the accident, i.e. the dynamics of the initial crash (maxi pile-up), the type of explosion (BLEVE), the place where the accident occurred (tunnel), the number of persons initially involved (over 50), and the rescue operations (immediate on-the-spot assistance), the following considerations can be made:

• the maxi pile-up in the tunnel blocked a considerable number of vehicles (22) and persons (over 50) in a closed and poorly illuminated environment

• all the passengers remained either near or inside their vehicles waiting for the traffic to move again. Nobody thought of getting out of the tunnel because it was raining. Nobody attempted to stop following traffic

• about half-way along the line of vehicles blocked in the tunnel there was a tank-truck carrying 2,500 litres of LPG

• people became aware of the imminent risk only after the first explosion, with the ignition of the cloud of gas and the sight of flames investing the upper part of the coach

• the ignition of the gas, which all witnesses described as a rapid gust of very hot wind, caused burns in exposed parts of the body (face and hands) in all persons (fourteen) some distance from the fire itself, without causing any damage to vehicles. Fortunately, as the day was cold and wet, people were wearing heavy clothing (Figs 9a, b, c, d)

Fig. 9a - Partial- and full-thickness burns Fig. 9b - Partial- and full-thickness burns.

Fig. 9c - Partial- and full-thickness burns. Fig. 9d - Partial- and full-thickness burns.

• the two most seriously burned persons were the drivers of the coach and the tank-truck, who were very near the fire. The coach-driver was the last person to leave the vehicle, having tried first to help the passengers; the tank-truck driver, having at first attempted to stop the traffic, then saw the coach approaching and skidding, at which point he moved off towards the exit, but fell down with his clothes aflame

• after the first explosion everybody had enough time to escape from the tunnel through one of the two exits, at a maximum distance of some 70 metres from the centre of the tunnel

• the BLEVE, in all its violence, did not occur for a further 6-7 minutes (all reports agree on this), and the physical damage testifies to its effects (five bodies were totally carbonized). The damage to the tunnel was immense. The shock wave sent the smoke through the connecting passageway into the adjacent tunnel, which was totally blackened. Everything in the tunnel where the BLEVE occurred was completely destroyed. It should be borne in mind that one kg of liquid gas produces 500 1 of gas which, mixed with air, becomes highly explosive

• the on-the-spot rescue work was fortuitous but also rapid and effective. A police patrol on duty on the motorway gave the first alarm by radio to its headquarters and to the emergency medical service, which in turn alerted the Burns Centre and two other hospitals in Palermo. The emergency services were also alerted by motorists with mobile phones

• the less seriously injured persons were taken by passing cars to Carini hospital (4 km) and to the nearest hospitals in Palermo. This accounts for some twelve persons who were medicated and discharged

• the general feeling of human sympathy enabled all the injured persons to use police telephones or mobile phones belonging to members of the general public in order to contact family and friends

• the first four ambulances arrived in loco after 35 minutes, with ten more after 40-45 minutes. Their arrival was facilitated by the fact that they were stationed on the city outskirts near the motorway

• sixteen patients were taken to the Palermo Burns Centre. None had infusion therapy during the transfer

• the first more serious patients arrived at the Burns Centre 70 minutes after the accident, and the last more than two hours later. After 2 hours 30 minutes the seriously injured patients were receiving infusion therapy and blood tests had been performed

• the Burns Centre was able to deal with the impact of the sudden arrival of this large number of patients thanks to an emergency plan tested on several occasions and also because of the state of preparedness for disaster emergencies in all the medical and nursing team.

1. Pugh R.W.: Quantify BLEVE hazards. Chem. Eng. Prog., 87: 66-72, 1991.
2. Arturson G.: The Los Alfaques disaster: a boiling-liquid, expandingvapour explosion. Burns, 7: 233-51, 1981.
3. Arturson G.: The tragedy of San Juanico. The most severe LPG 10. disaster in history. Burns, 13: 87-102, 1987.
4. Johansson 0., Wormuth D.W.: BLEVE ! The tragedy of San Juanico. In: "First", a journal of the Skandia International Insurance Corporation, Stockholm, Sweden: 2-23, 1985.
5. Masellis M.: Management of mass burn casualties in disasters. Ann.Medit. Burns Club, 1: 155-9, 1988.
6. Masellis M.: Thermal agent disaster and fire disaster: definition, damage assessment and relief operations. In: Masellis M., Gunn S.W.A. (eds), "The Management of Mass Burn Casualties and Fire Disasters: Proceedings of the First International Conference on Burns and Fire Disasters". Kluwer Academic Publishers, Dordrecht/Boston/London: 7-12, 1992.
7. Masellis M., Ferrara M.M., Gunn S.W.A.: Immediate assistance and first aid on the spot in fire disasters. Education of the public and self-sufficiency training. Ann. Medit. Burns Club, 5: 200-6, 1992.
8. Magliacani G.: Mass burns rescue operations: organization and medicosurgical therapies. Ann. Medit. Burns Club, 2: 199-204, 1989.
9. Dioguardi D., Brienza E., Altacera M.: The role of information sciences in the management of disasters. Ann. Medit Burns Club, 1: 165-7, 1988.
10. Mackie D.P., Konig H.M.: Fate of mass burn casualties: implications for disaster planning. Burns, 16: 203-6, 1990.
11. Hemdon D.N.: A survey of primary aid response to the Bashkir train gas pipeline disaster. Burns, 16: 323-4, 1990.

12. Masellis M., Di Stefano F., Colletti P., Pirillo E.: Mass burns in Sicily? Logistic welfare organization in relation to the regional Burn Centres and to the operating of military sanitary structures. Riv. Ital. Chir. Plast., 13: 174-85, 1981.