Toasty Talent: The 2016 Berlin (Germany) Train Fire
Background
Berlin is a city of 3.67 million people (as of 2024) in the far east of Germany, 125km/77.5mi east-northeast of Magdeburg and 146km/91mi north of Leipzig (both measurements in linear distance). The German capital is its own federal state, being entirely surrounded by but separate from the state of Brandenburg.
Being both the capital and largest city of Germany (by a margin of almost 2 million people) creates a high demand for transportation, leading to Berlin being filled with an expansive network of tramlines, subways, commuter- and regular rail lines alongside the road network and several airports. One of the city’s main rail lines is the Berlin Küstrin-Kietz rail line, a partially double-tracked electrified rail line from Berlin Ostbahnhof (“Berlin Eastern Station) to Polish border at Küstrin. The line opened in 1867 as the Preußische Ostbahn (Prussian Eastern Rail Line), being cut down to its current length after World War 2. The double-tracked line is accompanied from Berlin to Strausberg (28km/17mi) by a partially double-tracked line of the Berlin S-Bahn (suburban commuter rail), giving it a triple- to quad-tracked appearance. Nowadays the line is used exclusively for regional passenger service, with no freight traffic and the last cross-border service (the track does continue into Poland) being discontinued in 2009. The top speed is set at just 120kph/75mph.
The Train Involved
DPN 5179 was a regional passenger service provided by the NEB, running from Berlin-Lichtenberg to Küstrin/Kostrzyn on the Polish border. The company used a rented Class 643, numbered in their own system as the VT 736, a three-car diesel multiple-unit (DMU) made by Bombardier as the “Talent” between 1996 and 2008. Each three-car Talent measures 49.36m/162ft in length at an empty weight of 102 metric tons. The train was powered by two inline-six diesel engines providing a combined 630kW for a top speed of 120kph/75mph. The train had space for 156 seated and 160 standing passengers in a two-class configuration, and was carrying 150 passengers at the time of the accident.
The Accident
The train involved in the accident performed an earlier connection between Berlin and Küstrin (DPN 5177) around noon on the 6th of January 2016 with a double-traction, combining two three-car Talent trains. The trip was interrupted at 5:59pm when the onboard system reported to the driver that a train-rupture had occurred, meaning the system claimed the rear train had torn off at the coupler. This hadn’t actually taken place, but the issue led to the train actually being split up and the leading unit (the one which would have the accident) being sent to continue the trip by itself. The train continued its trip to Küstrin and back to Berlin without further issue, and departed eastbound again with an eight minute delay at 7:43pm, now as DPN 5179.
A few minutes later the train was approaching Berlin-Kaulsdorf station at 100kph/62mph when the onboard system reported a “timeout”-error (the report does not clarify what this means). At this point the train triggered an automatic stop, coming to a stop just before 8:00pm as the train autonomously shut down the engines. The driver radioed his dispatcher to report the unexplained emergency stop right when he was notified by passengers that the rear car was filling with smoke. He thus added a fire alarm to his report before using the emergency override to open the doors and have the train evacuated under the guidance of the onboard conductor. The passengers were guided to a nearby level crossing to get them off the rail line, before the conductor took a small group to the Wuhletal station 400m/1312ft up the line, presumably to find shelter from the freezing conditions.
The fire department arrived 15 minutes after the train came to a stop and found two of the train’s cars engulfed in flames, with the fire also having spread to the rear of the leading car. Thanks to the quick reactions by the crew all passengers had left the train without injuries before the fire spread, allowing the responders to focus on fighting the fire. They managed to extinguish the flames by 11pm, before the fire could completely consume the leading car too. The train was obviously a full loss, but everyone onboard escaped with nothing worse than cold feet from the snowy ground.
Aftermath
The federal police confiscated the charred wreck and interviewed the passengers the morning after the fire, initializing proceedings against unknown individuals on the suspicion of arson. The report openly states that the fire left them with an unusual situation, as a fire’s origin/cause is usually indicated by scorch marks, melted plastics or “burn notches” (localized missing material). But in this case the fire had so completely gutted the train car it started from that damage from the initial ignition and from the later, prolonged fire couldn’t be told apart. The investigation thus had to rely on witness statements, a cellphone video (when the fire was already burning) and what indications they had for things that DID NOT cause the fire to narrow down the cause. This process step by step ruled out arson, an engine fire, an exhaust fire or a fire in the air conditioning. The fire also started above the floor, ruling out the suspension and axle bearings. The witness statements pointed the investigators towards a pair of electrical cabinets near the rear driver’s cab, but the fire had left little beyond the bare metal framework of the train car and its interior, including the cabinets. The investigation consulted fire-experts from the TÜV Süd, who narrowed down the suspected origin of the fire to one of the cabinets, simply by checking which one had more power flowing through it and thus the higher potential to cause a fire in case of a malfunction. Fortunately the fire department had managed to save enough of the train’s forward end that the investigation had an identical set of intact electrical cabinets to compare things to.
The suspected cabinet (simply labeled “cabinet 1”) contained equipment in its upper section that dealt with the braking system and the general control system, while the parts below it were responsible for the horn and the coupling-system, including a heating system that kept the coupler operational in freezing conditions. The investigation concluded that the developing fire destroyed the components for the braking-system, triggering the emergency stop, while the sudden “disappearance” of the braking-system made the general control system shut down the engines as a safety measure to avoid further operation of the train.
The most likely culprit for starting the fire were two contactors in the lower part of the cabinet (6K4 and 7K1), which dealt with the coupler heating. Contactors are, in simplified terms, basically beefed up relays to carry higher loads. 6K4 and 7K1 carried the highest electrical load and thus had the ability to heat up the most. They were mounted on a pivoting frame which could swing out to allow access to the wiring behind them.
An examination of the type’s wiring diagrams and the documentation for the contactors used revealed that both contactors faced higher amperage than they were meant to, being “overloaded” by up to 60% in the case of 6K4 (9.6A instead of the intended 6A). Testing in an identical train during operation revealed regular loads of up to 8.5A. The investigators thus contacted the manufacturer (ABB) for input, who claimed that the components could withstand up to 16A if needed without failing. They explained that even at prolonged exposure to 16 ampere the contactors would continue to function without heating up enough to spark a fire. The manufacturer did note that they can’t guarantee that materials around the components (lubricants, cleaning liquids) wouldn’t be ignited. They also handed over paperwork showing that the contactors were meant to be mounted at least 5mm apart horizontally, which the contactors on the train weren’t (they touched). However, this was allegedly required simply because their operating temperature would lead to brown discoloration where they touch which, while harmless, would “unsettle” customers.
The investigators still went to Bombardier Transportation (who made the train) and asked them about the placement, being told that they understood the instructions as only being required for a special “uprated” version of the contactors to prolong their service life, thus leading to Bombardier deciding that they could mount the contactors in direct contact. In return the components would be swapped out every 8 years in case there was increased wear. The Class 643 is meant to undergo a major inspection after 8 years or no more than 1 Million kilometers (621 thousand miles), a mark the accident train had somewhat overshot by clocking 1.47 Million kilometers (913 thousand miles) by the time it came in for its first major inspection. The NEB had outsourced the inspection to a third party provider and handed a task-list to the investigation which, crucially, made no mention of replacing the contactors. The manufacturer’s inspection-guideline, in contrast, did require a visual examination every three months and a full replacement after eight years. The train caught fire at 11 years old and, as far as the paper trail showed, had never had its contactors replaced. Unfortunately the visual inspections, which were signed off on, involved no documentation like photos, and due to the complete destruction of the train car it’s unclear what condition the contactors were in at the time of the fire.
The investigators from the TÜV suggested that the fire likely started inside the cabinet when the worn and improperly mounted contractors overheated, being exposed to more than five ampere (required for sufficient heat) but less than 10 ampere (which would trip an emergency shutoff). The heat from the contactors then set insulation of the nearby wires and/or the pivoting frames ablaze, starting the fire. The flames remained localized within the cabinet for a short time, destroying the components for the braking- and general control system which triggered the alerts, emergency stop and autonomous shutdown. In the meantime the fire broke out of the cabinet but remained behind interior panels before it eventually melted the aluminum roof skin, at which point the improved oxygen supply increased the fire. Only now did the flames break through the ceiling panels, entering the passenger area and the rear driver’s cab. At this point the train was already stopped and evacuated, so nobody was injured as fire started falling out of the ceiling.
Lastly, the investigator from the TÜV narrowed down the most likely cause to contactor 6K4, which operates on a more powerful circuit than its neighbor. The degree of destruction made definite proof impossible, but the investigator suspects that a pulsating energization, which could be caused by vehicle motions “connecting and disconnecting” a lose connector/cable, caused the high thermal load which led to the fire. The report notes that prolonged heat exposure (over several years) wouldn’t just discolor the plastic but also reduce its ability to transfer/dissipate heat, meaning the older the contactor got the more it would heat up in the same situation. 6K4 is exposed to as much as 200 watts, while its neighbor only has to deal with 150 watts, making 6K4, more specifically its contacts, the likely origin of the fire. According to the report continued wear on the contacts increases the resistance provided by the contacts, and higher electrical resistance leads to increased warmth being generated. This theory was backed up by comparing the train’s surviving contactors with brand new counterparts in a laboratory. The contactors which supposedly started the fire were 3 years overdue for replacement, further supporting the theory.
It’s unclear what exactly caused the false train-rupture alert early in the day, but it could have been the first sign of faulty cables and/or connections, which then worsened over the day as the train bumbled along its route, shaking about and with that shaking all the cables, including a fatal lose one which ended up being the pulsating energization.
The NEB reacted by replacing the two contactors on their whole Talent-fleet and changing the wiring of the cabinet so the highest loads were distributed on more contacts. Inspection- or installation-rules, however, remained unchanged. It was likely assumed that the maintenance department had “slipped up”, and that the fire wouldn’t have started on contactors no older than 8 years. There is no evidence that specifically 8 years old contactors were tested to observe their behavior in the given scenario.
_______________________________________________________________
I have chosen to stay off Medium’s monetization offers to keep these stories as accessible as possible, but you can support me with a small tip via “Buy me a coffee” if you feel like it.
_______________________________________________________________
A kind reader is posting the installments on reddit for me, I cannot interact with you there but I will read the feedback and corrections. You can find the post right here.
