Disastrous Delivery: The 2020 Drackenstedt (Germany) Train Derailment
Background
Drackenstedt is a town of 364 people (as of 2018) in central Germany, located in the federal state of Saxony-Anhalt 19km/12mi west of Magdeburg and 85km/53mi southwest of Wolfsburg (both measurements in linear distance).
The town lies on the Braunschweig-Magdeburg rail line, an 83km/52mi electrified dual-track main line which is considered one of the most important east-west-lines between Hannover and Berlin. The line opened in 1872, quickly becoming one of the most important east-west lines, eventually also serving as a transit-line between West-Germany and West-Berlin after WW2. Nowadays the line sees a lot of freight traffic but also regional and long-distance passenger services at speeds of up to 160kph/99mph.
The Train Involved
DbZ 79653 was a transfer-service from Salzgitter to Blankenburg (Harz), conducted by the VGT (Vorbereitungsgesellschaft Transporttechnik/Preparation-Company for Transportation Technology), a private company specializing in testing new trains and educating staff for the work with them. The train consisted of a Class 214 locomotive owned by Alstom (a French train- and railway equipment manufacturer) towing a pair of brand new NS (Dutch national railway) ICNG-units. The Dutch railway uses the abbreviation ICNG (“InterCity New Generation”) in place of a number to classify the type. The ICNG train isn’t yet permitted to operate in Germany and was being transported to Blankenburg for some undefined testing procedures at that town’s station.
The Class 214 is a modernized version of the old DB (German national railway) Class V100, a four-axle multipurpose diesel locomotive, introduced by Alstom in 2008. The modernization uses little beyond the frame and bogies of the 1950s-developed V100, but retains their distinct shape with the centered driver’s cab and lower hoods which the driver can look over. The resulting locomotive is 12.3m/40ft long at 61 metric tons and can reach a top speed of 100kph/62mph based on a 970kW power output. The locomotive involved in the accident (214 006) had been built in 1963 and converted in 2010. The locomotive was fitted with a coupler-adapter to be able to connect to the ICNG’s Scharfenberg-type coupler.
The ICNG is a five- or eight-car electric multiple unit (EMU) launched by Coradia as part of the “Stream”-family of trains in 2018 and is intended for express passenger services in the Netherlands, with some units being fitted with equipment required to operate routes into Belgium and Germany. The two units involved in the accident (3105 and 3109) were five-car versions. Each five-car ICNG measures 110m/361ft in length at 199 metric tons and can carry up to 256 passengers at as much as 200kph/124mph in a two-class configuration. The two trains were completely unmanned during the trip and acted essentially as if they were just passenger cars, with their brakes being controlled by the locomotive up ahead. The trains are allowed to be towed at up to 100kph/62mph.
The Accident
The VGT’s driver is hooking his locomotive up to the pair of new EMUs at Alstom’s Salzgitter facility early in the morning of the 16th of October 2020. His locomotive is fitted with a coupler-adapter to accept the forward EMU’s Scharfenberg-type coupler (a type not used on regular locomotives), but the air lines for the brakes connect directly between the forward EMU and the locomotive. A full, standard test of the brake-system once everything is connected is passed without incident before the train leaves the facility a 8:19am, stopping at nearby Salzgitter-Immendorf West station without issues. Another 2 stops and 13 decelerations are performed without incident by the time the train reaches the pre-signal for the entry-signal of Dreileben-Drackenstedt station, which shows “expect stop”. The train is meant to stop in a siding at the station.
The driver, obviously planning to obey the instruction, gradually opens the brake-valve to start slowing the train, but is met with what he later describes as a marginal deceleration. The train now breaches the reduced speed limit behind the pre-signal, causing the PZB (punctual train control)-system to trigger an emergency stop. But even the fully opened valves fail to deliver any meaningful deceleration. The train thus passes the entry-signal at 10:10am, travelling at 88kph/55mph where it was meant to stop. An emergency call from the train to the dispatcher is recorded at that point, but with no recorded statement from the driver. The train races through an open level crossing and enters the station as another “empty” emergency call is placed before the driver, in a third call 4 seconds after the second, speaks up (from the report, translated):
Driver: Hello?
Dispatcher: Where are you?
Driver: Yes sorry I’m at…uhh…
Dispatcher: You can’t keep placing emergency calls and keep driving. Where are you?
Driver: I’m…uh…at Dreileben-Drackenstedt.
Dispatcher: Why don’t you stop there? Why do you keep making emergency calls out of boredom?
Driver: I triggered an emergency stop and I’m about to hit a buffer. Uh I’m right ahead of a buffer stop.
Dispatcher: So?
Driver: Shit, now, yes now everything’s over (unintelligible)
Dispatcher: Well thanks!
Driver: I can’t [loud noise]. Ah shit…[loud noise]
The buffer stop the driver is referring to belongs to a set of catch-points at the end of the station, a special kind of points designed to keep unpermitted trains off main lines. In this case, they direct traffic that has no permission to leave the station into a buffer stop at the end of a siding barely 30m/100ft long.
The locomotive passes the exit-signal just 90 seconds after entering the station, still at 41kph/25mph, and is directed into the short siding by the catch points. It obliterates the buffer stop at the end of the dead-end siding, being propelled slightly upward by the debris before its front end digs into the sand. It’s rear-ended by the leading EMU at the same time, causing the locomotive to flip over onto its roof, crushing the driver’s cab down to about the height of the hoods to either side of it. The leading end of the leading EMU gets stuck on the front end of the upturned locomotive despite the force of the second EMU pushing from behind, leaving it with several axles derailed. Torn fuel and oil lines start a small fire on the destroyed locomotive as the wreckage comes to a rest. The driver is severely injured in the derailment, but manages to climb out of the wreckage before responders even arrive at the site.
Aftermath
The direct cause was fairly obvious, the train’s brakes had failed to stop the train, leaving it to derail on the catch-points as it had no permission to depart the station. This couldn’t have been the driver forgetting or neglecting to apply the brakes, as the PZB-system would’ve brought the train to a stop independently of driver inputs if that had been the case. The investigators started their examination of the day’s events with the train’s pre-departure preparations early in the morning. The documentation showed that the two EMUs had been shunted to the intended departure-track at Alstom’s facility around 6:30am that morning, at which point their brakes were checked to ensure functionality. The locomotive arrived in the same track by 7:00am, being coupled to the leading EMU with the coupler-adapter. The driver stated that he briefly opened the valves for the main air reservoir line (HBL) and main pneumatic line (HLL) to clear debris from the hoses before connecting everything.
The driver then used the locomotive’s remote control (a common feature in shunting locomotives) to perform a full brake-test, using the guidelines for such a test on locomotive-pulled passenger trains (as the EMUs were running “cold”, acting like passenger cars). He entered the leading EMU (number 3105) and used a pressure gauge found inside it to ensure that the brakes released as intended before applying the brakes. Guidelines demanded that the driver now check on the rearmost brake of the train if they were all applied, in this case the driver checked on each brake individually. He then returned to the cab of his locomotive and checked proper operation of the brake from the regular controls installed there. At last he filled out the form for train-weight and brake-force, slightly miscalculating the braking capability of the train (arriving at a braking-percentage of 113 instead of 110). However, the report notes that the route only required 90, so the capability was sufficient either way.
The EMUs’ onboard batteries were turned on by 8:10am, which also activated the measuring equipment installed aboard EMU 3105 for the scheduled tests. The main reservoir line was logged at 4.3bar/62psi at that point, when it should’ve been at 8.5bar/123psi. The train then departed the facility at 8:19am. A total of 12 brake-applications were successful over the following 80.7km/50.1mi journey before the train reached the pre-signal outside Dreileben-Drackenstedt station, which showed “expect stop”. Here, the driver was suddenly met with insufficient deceleration, an issue not solved by the triggered automatic stop which dumped air pressure from the main pneumatic line. An automatic emergency stop, triggered by the PZB 22 seconds later, was without effect as air pressure had already been dumped from the locomotive. The train thus proceeded through the main entrance-signal and an open level crossing, still travelling at 70kph/43.5mph. It’s pure luck that nobody was using the crossing at that moment. A few seconds later the train reached the catch-points and derailed.
With that information on hand the investigators first examined the compressor aboard the locomotive. The installed unit was able to provide 2500l/660gal of air per minute at a pressure of 8.5–10bar/123–145psi. The pressure in the HBL was continuously monitored by the locomotive, with the compressor designed to kick in at a pressure of at least 5bar/72psi. A lower pressure would’ve automatically caused the pressure in the HLL to be reduced, applying the brakes. This meant that the compressor worked as intended, as the pressure logged by the EMU never even reached 5bar, meaning the brakes shouldn’t have been released. Attention thus turned to the valves on the back of the locomotive, which were found with some impact-damage.
The valve-handle for the HBL was found in the “closed” position, one for the HLL was set to “open”. A second HLL-handle was broken off and, lastly, the second HBL-handle was in between the open and closed positions. The left hand HBL and HLL-handles’ mounting plate was partially torn off the locomotive and impact-marks on the left hand HBL-handle indicated that it was exposed to forces in the direction that would set it to “open”. The extend of the damages made it impossible to confirm whether or not the HBL-handle had been opened during the pre-departure procedures, but the fact that the HBL of the leading EMU wasn’t pressurized properly implies that it had erroneously been closed.
The report theorizes that the brakes on the EMUs were released during the brake-tests, and that the HBL-valve on the locomotive was then erroneously closed. In normal operation a brake-application would reduce air pressure in the HLL, allowing air pressure from the HBL to “overcome” the HLL’s pressure to apply the brakes. Releasing the brakes again vented the air to the outside. As the locomotive (the only part of the train with a working compressor) couldn’t refill the HBL accordingly pressure gradually dropped from brake-applications and smaller leaks, leaving it at approximately 0.6bar/8.7psi by the time the train approached the entrance-pre-signal to Dreileben-Drackenstedt station. The driver had already released more pressure than usual from the HLL during the last stop, but apparently didn’t notice this oddity.
The approach to Dreileben-Drackenstedt station involved a complete dump of air pressure from the HLL, but by that point the HBL only had 0.3bar/4.3psi left in it, so the brake cylinder pressure was neglectable to nonexistent. Thus, the brakes on the EMUs didn’t work. The brakes on the locomotive worked as intended, but its brakes alone had no chance to slow the train down when a total of 398 metric tons without propulsion but also without any brakes were pushing from behind.
The report concludes that the main cause of the crash was the non-existence of continuity between the locomotive’s and the EMUs’ HBL, leaving the brakes aboard the EMUs increasingly non-functional as the EMUs’ HBL couldn’t be refilled while the locomotive’s onboard system could only observe pressure in the locomotive’s HBL, which was fine as the valve was closed, keeping the air from escaping uncontrollably the way a leak or torn hose would cause it to. The condition of the locomotive after the crash didn’t allow for certainty in the theory, but it was assumed that the valve-handle on the HBL wasn’t fully opened ahead of the train’s departure from Alstom’s facility.
Apart from the “causal factor” noted above the report also calls out a systemic factor, namely the insufficient control measures regarding the effectiveness of the brake” as a failure in the VGT’s safety management systems. Apparently Alstom had not informed VGT about the function of the EMU’s brakes, only providing a German and an English manual for the train which didn’t mention the function of the brake-system, and VGT also hadn’t inquired about it. The VGT argued that they knew braking systems like that of the towed EMUs because it’s similar to the brakes of other Alstom-trains built over the last 10 years for Germany. And since they had tasked a highly experienced employee with the trip they had figured that everything was going to be alright since he had experience with those systems.
Since the EMUs weren’t intended for Germany the VGT had been required to have a “harmlessness certificate” created that guaranteed the safety of the EMU moving on German rails, even if towed. The certificate, fatally, treated tows and test-drives as one. The report notes that a proper treatment of the EMU as a rail-vehicle not intended (or registered) for service would’ve involved several repeated tests of the brake-system (in accordance with guidelines the VGT didn’t seem to have on hand) and observations of the EMU’s HBL-pressure, where the lack of air pressure would have likely become obvious. Communication between the investigators and the VGT revealed that they had planned to have the EMUs’ HBL pressurized between 5 and 10 bar/72–145psi to ensure proper controllability of the EMU’s brakes. As the data-logger showed after the crash the pressure was down to 3.2bar/46psi by the time they stopped at Salzgitter-Immendorf West station not even 15 minutes into the trip. Proper observation of the HBL-pressure would have shown the brake-defect at this point, before it turned into a brake-failure. However, there was no way to monitor the EMUs’ HBL-pressure from the towing locomotive. And the certificate, treating test-drives and tow-drives the same, didn’t note this risk. Because during test-drives there is someone aboard the EMU. The investigation said that this failure of proper risk-assessment, part of an insufficient safety management, allowed the faulty setup of the train’s brakes to turn into brake-failure.
_______________________________________________________________
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.
