The Landrücken Tunnel is Germany’s longest railway tunnel at 10.8km/35300ft, allowing the Hanover-Würzburg high speed line to pass straight under a mountain range in the federal state of Hesse in the southwest of Germany. The southern portal can be found 1.5km/5000ft north of the town of Mottgers, right on the border between Hesse and Bavaria, while the northern portal (which is more important for us) lies just 16.5km/10.25mi south of the city of Fulda (all distances measured in linear distance).
Opened in Mai 1988 the two-tracked electrified route through the tunnel not only cuts a major detour around the “Landrücken” mountain range but also offers a much straighter track, allowing speeds of up to 250kph/160kph. Technologically the tunnel is up to date, offering modern signaling-systems with block-sections (meaning trains getting too close to one-another get auto-stopped), radio communications, cell service, two emergency exits and even reception of normal everyday FM radio programs. It also holds sets of points in two places, allowing trains to switch between the tracks if needed.
The trains involved
Passing through the tunnel first was ICE 782 from Munich to Hamburg run by ICE 2 unit number 214 (named “Hamm (Westphalia)”, a German city) with the motor car 402 014 leading. The ICE 2 is the second generation of the Intercity Express, Germany’s high speed passenger train. Introduced in 1996 these trains, consisting of a motor car without seats, five passenger cars, an onboard bistro (called the Bordrestaurant) and a control car with seats, can carry 368 passengers in a two-class configuration at 205m/673ft long. Weighting 412 metric tons empty they are capable of reaching 280kph/174mph, plenty for most German high speed rail lines. They can be distinguished from their predecessor by having only one motor car and having a restaurant car with a flat roof while the predecessor’s bistro car’s roof was raised 450mm/17.5in higher than the rest of the train. They are also much shorter than their predecessor, consisting of 8 cars instead of 14.
Travelling in the opposite direction was ICE 885 from Hamburg to Munich. On the day of the accident the connection was provided by ICE 1 number 111 (named Nuremberg), a 14-unit first generation ICE measuring 358m/1175ft in length. Entering service in March 1991 TZ (“Triebzug”, German for “railcar” or “multiple unit”) 111 was capable of carrying 703 passengers in two classes at up to 250kph/174mph, but on the fateful day only carried 148 people including the crew. Following the leading motor car were four first class passenger cars, followed by the restaurant car, seven second class cars and the rear motor car (401 011)
In the evening of the 26th of April 2008 ICE 782 from Munich to Hamburg is racing through the Landrücken Tunnel on it’s way north, travelling at 220kph/137mph. As the train approaches the northern portal the driver spots an unidentified obstacle on his track. He triggers an emergency stop at 8:59:52pm. The train strikes the obstacle, later identified as a single sheep, regardless and takes 1615m/1mi to come to a stop right outside the portal. The driver exits his motor car and examines the front of his train, finding blood and some mostly cosmetic damage, along with a disabled LZB-antenna.
LZB stands for “Lineare Zugbeeinflussung” (“Linear Train Control”), a system in use in Germany, Austria and some Spanish rail lines that tracks the train’s position with antennas in the track and on the train, monitoring the surrounding signals and controlling/limiting speed. The system can tell the train driver what speed to go at any moment, show signals that are yet to come into sight (calculating braking-distances). With this system block-zones don’t have to extend for higher speeds, allowing a higher frequency of trains going in the same direction on the same track without compromising safety.
A speedometer on a train with LZB displays 3 speeds. The yellow indicator shows the current speed, the red marker the speed the train is permitted to travel at most. The red number below the dial shows how fast the train is meant to go at that moment. If the system calculates a stop the latter figure would sink to 0.
With the LZB-system on ICE 782 knocked out from the impact with the sheep the driver is permitted to restart the train after 2 minutes and 3 seconds stopped, now at no more than 160kph/99mph. This is the maximum speed trains without LZB can travel at in Germany, keeping braking-distances short enough for safe operation. No-one aboard the train got injured, and the damage was presumably repaired within the next few hours or days.
At the same time as ICE 782 starts driving again it’s opposite train, ICE 885 from Hamburg to Munich provided by TZ 111 led by motor car 401 511 is approaching the tunnel from the north. Having just left Fulda station the 800 metric ton train is headed slightly uphill, topping out at 210kph/130kph as it leaves the Bornheck Tunnel about 1km/3280ft north of the Landrücken Tunnel. At 9:04:40pm, just 7m/23ft inside the Landrücken Tunnel the data-logger on TZ 111 registers the LZB-system going offline, most likely caused by hitting a small group of sheep. The train strikes another, larger group 15m/49ft further into the tunnel. Reduced grip for the motor car’s leading axle leads the data-logger to register a momentarily increased speed of 215kph/134mph for just 35m/115ft. The dirty/blocked tracks cause the leading axle to slip out of the track 60m/197ft inside the tunnel. This is later proven by scratches on the tracks and sleepers.
The driver triggers an emergency stop, the train dumps air pressure and starts decelerating at 9:04:45pm, five seconds after the train struck the first sheep. A recognition-reaction-effect time (the time from seeing something to the taken action taking effect) of five seconds is considered perfectly fine in the report. Immediately after the stop is initiated the shaking of the motor car ges violent enough to throw the driver out of his seat. The train keeps going relatively straight for 700m/2297ft, gradually loosing speed. At that point it reaches the first set of points (No 602), which is set up to allow diversion for trains in the opposite direction. The derailed axle strikes a wheel guide and is deflected sharply to the right, derailing the second axle at 174kph/108mph. The nose of the train strikes the wall of the tunnel, causing significant damage to the motor car before being deflected back towards the track. The forces of the impact also move the entire, connected and welded-in set of points to the left. Crossing over the variable part of the points the forward axles are forced back to the left, leading them almost back to being as close to the rails as they were before. The damage the impact and derailment caused to the track is severe, the report lists “barely attached rails” and destroyed sleepers in the next 200m/656ft, saying there was “objectively no track left”. The report fails to explain how the first two passenger cars stayed in line, with only the third to last passenger cars and rear motor car derailing and leaving the (former) track. After the set of points the nose of the train moves to the right again, scraping along the wall (destroying power lines, telecommunication-systems and the wall itself) before the train finally stops at at 9:05:11pm, 107m/150ft inside the tunnel.
1 crew member and 21 passengers suffer severe injuries in the derailment, caused by the violent motions of the train and loose baggage moving around the interior, while 4 crew members in the restaurant car and 13 passengers suffer minor injuries.
Immediately after the train comes to a stop the driver manages to issue a route block command, meaning no train can and will use the oncoming track. This avoids the risk of another train hitting the derailed train cars, something that has happened before. The driver then receives a status report from the conductor and stewards, saying all passengers are capable of leaving the train on their own/with light assistance. The crew in the train car collects the passengers into groups, leaving most of the luggage behind, before leaving the train and making their way along the side of the tunnel towards the northern portal. The train has kicked up a high amount of dust, severely limiting sight and making the signage hard to read. Some passengers reportedly think something was on fire, aided by a smell of burned rubber (friction heat).
By the time they reach the northern portal responders are arriving at a designated rescue field outside the portal and can tend to the injured passengers.
The DB’s (German national railway) emergency response team arrives at the portal at 9:30pm, after being accidentally sent to the side-portal of an escape tunnel first. Most passengers and the crew of the train were treated by medical staff on site and then taken to Mittelkalbach (1.8km/1.1mi linear distance away) by bus, where they could be hosted, receive their luggage once it was recovered and continue their journeys. Only 3 people required hospitalization due to fractured bones.
The recovery-train stationed at Fulda main station was alerted at 9:33pm and arrived at the site 25 minutes later, it’s opposite partner from Würzburg reached the site (inside the tunnel) at 0:44am the following day. The delayed deployment of the latter was later blamed on the emergency manager on site (at the northern portal) underestimating the situation inside the tunnel. Investigators of course noticed the large amount of dead sheep scattered across the track near the portal, the report even notes an “extreme smell of decay” in the days following the accident making the investigation and recovery more difficult. Following measuring of selected sheep and counting of the remains it is decided that striking 3.7 metric tons of sheep most likely caused the accident, this theory is essentially proven when no pre-existing defect (something not caused by the derailment) can be found on the train, in the data-logger or on the track. As such, the report lists three possible causes for the derailment of the leading motor car.
- Theory 1: Striking the sheep caused the motor car’s leading axles to lift up enough to derail (supported by the driver saying he felt the train car pitch up momentarily). Computer simulations later show that 9–13 sheep (equivalent of 0.9–1.3 metric tons, not even half of what’s found on the scene) can be enough of an obstacle to cause this. This is in contrast to testing done before the introduction of the ICE 1, which showed heavy damage but a nearly nonexistent chance of derailment. The main difference is found in the obstacle, during pre-introduction testing single, large animals were assumed as the obstacle (like deer), not a heard of relatively small animals. The chance of several dead animal’s remains mounting up under the train was also not taken into consideration.
- Theory 2: Components damaged during the collision with the herd of sheep became lodged under the train, along with several compressed, dead sheep. The resistance from this foreign object being pushed along by the forward bogie damaged the track and derailed the train, with the lift felt by the driver coming from momentary resistance when the tracks were damaged and the created object overrun. This theory is supported by a single cracked sleeper found ahead of where the train actually derailed.
- Theory 3: An unlucky combination of both. Since there is evidence for both theories and against neither, this is the answer the report settles on.
A few days after the accident the DB started to remove the train from the tunnel in sections, due to the damage caused by the derailment this had to be done through the southern portal. Most cars managed to roll on their own after being re-tracked, the rest (including the leading motor car) were towed out after being placed on temporary wheelsets.
The train cars were pulled to the nearby town of Mottgers, which does not have a station but does have sidings allowing for safe storage of the damaged train cars. The accident had caused 10.3 million Euros/12.5 million USD in damage (excluding the animals), and the tunnel only fully reopened in mid-June 2008 causing countless delays due to trains needing to take the old, longer and slower route or (with repairs advancing) only being able to use one track at a reduced speed.
The owner of the sheep initially finds himself the target of a criminal investigation, based on charges of dangerous interference with rail traffic and several cases of negligent cause of bodily harm. In February 2009 Fulda’s public prosecutor’s office ceases to pursue charges against him. He is not legally required to watch his sheep at all times, and the fences of the field the animals escaped from were up to the requirements. The official explanation is that unelashed dogs caused the sheep to escape their meadow in panic and run into the nearby tunnel. This sparks a renewed discussion about the safety of German high speed railway lines, as there were no fences to avoid something like this from happening. The DB points to French high speed lines, which are fenced in 2m/6.5 feet high and still suffer collisions with wildlife. The explanation is that fences can be passed by wildlife, which then can’t escape as easily as it could without fences. Furthermore, the DB claims that fenced in tracks would severely slow down rescue and recovery operations. They claim that their procedure, having experts find the areas with the highest risk of wildlife crossing a proposed railway line and planning with tunnels or bridges for wildlife to be sufficient. While this explanation is enough to escape criminal charges the DB has to give in eventually, and in late 2011 protective fences are installed at the northern portal. The southern portal remains relatively open, with only guardrails keeping vehicles from reaching the track.
The report closes with another few notes, criticizing poor visibility/recognizability of escape routes and emergency equipment in the trains (something several passengers complained about), advising improved training of response-teams (avoiding late deployment or deployment to incorrect locations) and recommending more regular training of local responders for train-accidents.
TZ 111 is initially split up after the accident, with the first class cars replacing their counterparts in TZ 173 which suffers an accident in Switzerland in April two days after this one. In 2013 the train is reassembled, the first class cars return as TZ 173’s original cars are repaired. Today it’s back in regular service with all it’s original cars, with no sign of the accident.