The Channel Tunnel Fire

This was an quite different fire as it broke out in an railroad tunnel. But as this tunnel is also one of the most used crossing points from the UK to the Continent about 50% of all vehicles use it.

The following report is based mainly on the Department of Environment, Transport and the Regions's Channel Tunnel Safety Report. Here the french text. contains the official report. is a report on safety in tunnels

The Channel Tunnel consists of 3 separate tubes burried about 40 meters under the ground of the Sea. It's lenght is about 50 Km of which 38 Km are below the water. The outer North- and South tubes of 7,6 meters each contain a single lane railroad line. The middle tube of 4,8 meters is reserved for maintenance and emergency. The outer tubes are connected to each other to equalize air pressures from the running trains. The emergency tunnel , 8 meters between each outer tube is always separated by airtight shelter doors every 375 meters and is overpressurized.

Freight and car trafic is separated on different trains. In passenger vehicle trains each car forms an airtight shelter in itself, the freight train cars each contain a single HGVon open platform cars. Drivers have a separate, supposedly airtight, saloon car just after the front train head.

20.42 (8.42 PM) on Monday, November 18, 1996

The freightrain #17539 leaves Sangatte and slowly winds it's way through the vast courts in front of the southern tube portal. Aboard it's 30 cars are 29 lorries (HGV's), 31 drivers and 2 railroad employes. It's already pitch dark outside.


Six minutes after departure an safety agent calls the trafic control room by phone and reports having seen 1-2 meters high flames emanating from a lorrie aboard the train. This information is immediately forwarded to the train control. 3 other persons also call in and report having seen flames. The subofficer Dave Wales from the Kent Firefighters, commanding the emergency center of the tunnel gets called to the manager in duty at the control room.

In the minutes following the entrance into the tunnel a serie of different warning systems in the tunnel and aboard the train gave alarm, of the six fixed fire detectors only 5 gave a "simple" alarm (an "unconfirmed" alarm). The fire detector systems on the tail trainhead only came on very late. Video systems transmitting their pictures directly to the control room confirmed an serious incident in one of the last cars of train #17539. Emergency action plans dictate to get the train out of the tunnel onto a special lane where firefighting and rescue actions can be attempted. So the machinist gets the order to carry on towards Folkestone. Meanwhile the Chunnel's radio frequencies got overloaden with alarm messages, even the phone line to the british side got "lost".


Alarm for the STTS (Service Tunnel Transportation System). vehicles #1 and #2 which are constantly equipped with firefighting equipment immediately engage into the middle tube. Both accelerate to 80 Km/h, the maximum speed and head for the middle of the tunnel following the emergency plans. At this moment the train is still running.

Two STTS in the middle emergency and maintenance tunnel. They are guided by an line under the floor.


The fire spread fast to the rear probably helped by the train running. At 20.58 it reaches the cables of the safety system and burns it, creating an "overspeed emergency breaking" override command. So the train driver is forced to slow down and working against the braking system as he searches for an exit door. Smoke entered the now slow running saloon car, even more smoke entered when the train conductor had to open the door to evaluate the situation. Finally the train stopped with the sallon car's door just opposite the exit #4134. It's shelter door lies 19 Km from France and 31 Km from Britain.

The train driver tries to activate the second emergency system which disconnects the saloon car and the train head from the rest of the freight train after which they should be driven outside to bring the passengers on safe ground. But the fire was already so hot it immediately burned parts of the catenary, cutting the power off only seconds after the stop. Immediately evacuation began in the dim lighting of the emergency lights. The cold-blooded and professional railroad employes managed to evacuate the 34 people inside the emergency tunnel although thick poisonous smoke immediately filled the tube and the fire quickly propagated towards the front. Luckily the burning HGV was quite at the opposite train end.

The smoke got driven from the french to the british side, largely due to the ram-air effect of trains still following. The nearest one was 4 kilometers behind and it's speed was governed by signals. If I suppose right (the official report isn't clear on this) at least 2 trains were following the burning train.


Both STTS vehicles arrive at #4134 half an hour after the fleeing people entered the safety tunnel. Several people showed signs of intoxication and they received first aid. An oncoming train was stopped in the northbound tube and transported the passengers out. 8 people were brought into hospitals.

Now firefighting could start. The aditional aeration system brought lots of fresh and clean air from the british side. The men advanced into the tunnel, helped by the air in their back. The front train head and the first cars were in good condition, the rest of the train was hidden in a curb.Soon they could see and hear the fire.


Eurotunnel Control alerted the Kent commanding post which sent 7 engines with 24 men. At the tunnel they had to switch into STTS 3 and 4.


The men arrive at the fire. The french firefighters already work from the next door at #4163, further to the back of the train. Remember, the #4134 is near the train head and was used to evacuate the passengers. The british forces decide to open the next door, #4201 where the center of the fire is expected. When they open the heavy door the high pressure in the safety tunnel pushes the air out with violence, clearing the area immediately around the door from heat and smoke. Temperatures stayed stable at the door but only a step further a furnace was burning. The heat and moistness were so high none of the firemen had ever encountered such conditions. Even the high quality protective equipment allowed only a 15 minutes duty cycle in what was later discribed as "Obstacle run in a sauna". The train only left a little more than a meter space to work in. Burning parts of loads fell off, concrete from the ceiling broke down as temperatures reached 1000 °C.

It was great luck that the sea ground in this place had no cracks or crevices where water could REALLY come through- apart a bit of seeping. Else the whole tunnel would have been flooded and without doubt this had cost some lives.

The concrete was completely destroyed and vanished leaving only the bare metal structure which lost it's strenght and bent down in the intense heat. Clearly the bare stone walls can be seen between the metal.

A complete commanding post is installed in the emergency tunnel. More than 200 oxygen bottles were in use, refills had to make a 80 Km long run. 6 firefighting engines were loaden aboard a train and brought to the site in the northern tube at 1.47 AM.

The completely destroyed rear trainhead still inside the Chunnel

Tuesday 9.00

The fire is extinguished although it took until 16.00 to declare the area safe again.


Heavy damage at 10 Lorries, 5 further got lightly damaged. Heavy damage to cars and the rear train head. Heavy damage at 2 kilometers lenght to the tunnel tube, some segments broke completely down. All technical systems had to be replaced including rails, catenary, all cables, some shelter doors, the water ducts and a survey station with all sensors.

Again heavy structural damage to the walls and railroad cars is visible

Investigations showed no signs of lack oxygen at the fire. The heat developed was calculated at 500 Megawatts, enough to melt the wheels to the rails. The fire was voluntary laid.

The tunnel was shut for 6 months until May 15, 1997. It caused 200 million damage--but no lives were lost. No major incident occured since then.

Although this incident was quite serious it proved the Chunnel to be by construction one of the safest railroad tunnels. Sure, problems occured during this fire but situations like this can't be predicted in every detail.

Investigations on tunnel fires answered some questions. If there is a burning HGV in the channel tunnel and we suppose there is a ventilation velocity of two meters per second (about 0.5 miles per hour). The model predicts that the fire will spread to the next HGV if or when the heat release rate of the fire exceeds 55MW. At 2.5m/s the required heat release rate increases to over 70MW. At 3m/s the fire is not likely to spread.

Professor Ulm of the Department of Civil and Environmental Engineering and two coauthors -- Professor Olivier Coussy of the Laboratoire Central des Ponts et Chaussees in France and Professor Zdenek Bazant at Northwestern University -- discovered that when mature dried concrete is exposed to extreme heat for long periods of time, the chemical bonds between the water molecules in the concrete break, destroying molecular bridges that bind together the various materials that make up concrete. As the water molecules are pulled out of the skeleton through dehydration, the concrete loses its cohesion and weakens, pushing pieces of the concrete off the tunnel walls in very thin layers resembling onion peel. This phenomenon, called spalling, can eventually work its way through the entire concrete ring lining a tunnel, layer by layer. "After the [Chunnel] fire, there were pieces of the spalled concrete on the tunnel floor. You could actually see the aggregates in the material in these thin slices," said Professor Ulm.

Last updated 03.01.2004