Slipped off: The 2020 Stonehaven (Scotland) Train Derailment

Max S
18 min readMar 20, 2022

--

Background

Stonehaven is a town of 11170 people (as of 2016) in northeast Scotland, located in the council-area of Aberdeenshire 25km/15.5mi south-southwest of Aberdeen and 123km/76mi north-northeast of Edinburgh (both measurements in linear distance).

The location of Stonehaven on the British Isles.

Stonehaven lies on the Dundee-Aberdeen line, a non-electrified largely double-tracked main line running along Scotland’s eastern coast. It is mostly used for passenger services, from regional trains up to long-distance overnight express trains connecting Scotland with London. Opening in sections between 1838 and 1883 it is an important part of Scottish railway infrastructure, with trains nowadays travelling at up to 161kph/100mph on the line.

The site of the accident seen from above in 2018. The train came from the northeast (top of the image).

The train involved

1T08 was a high speed passenger service scheduled to travel from Aberdeen to Glasgow, running under Abellio ScotRail’s new “Inter7City”-branding. The service was provided by a refurbished Intercity 125-set, a six-car diesel-powered high speed train. Introduced into service in 1976 the IC125 is a high speed diesel express train consisting of 2 British Rail Class 43 locomotives with a row of Mark 3 passenger cars between them. The Class 43 is a four-axle high speed diesel locomotive purposely developed for the then-new Intercity service, evident in its asymmetrical design with a streamlined and a “blunt”end, giving the IC125-sets the appearance of a standard multiple unit. Originally a small stripe on the vertical end also wore the same livery as the passenger cars, making the trains look more uniform. This was later extended to the whole locomotive’s livery matching the cars. Each Class 43 weights 70.25 metric tons at 17.79m/58ft in length and has a listed top speed of 200kph/125mph (giving the Intercity 125 its name) while 238kph/148mph have been reached in testing. Between 2004 and 2010 each Class 43 was fitted with a by a 65l/3967cubic inch MTU 16V4000 R41R sixteen cylinder diesel engine producing 1700kW/2280hp at 1500rpm. 197 units were made for British Rail between 1975 and 1982 and allocated to different regions before being later leased to different rail service providers.

A six-car IC125 like the one involved in the accident consists of the two locomotives (referred to as “power cars”) and 4 Mark 3 passenger cars (called “coaches” in the UK). Introduced in 1975 each Mark 3 passenger car measures 23m/75.6ft in length and can carry 74 (second class) or 48 (first class) passengers at up to 200kph/125mph. The cars have an all-steel construction and weight 33.6 metric tons each. At the time of the accident the train carried just nine people, being basically empty. The group was made up of the driver, a conductor, six random passengers and another conductor who used the train to shuttle to a different station where a train was waiting for her to work on.

On the day of the accident the unit running as 1T08 was led by Class 43 number 43140, with locomotive number 43030 trailing. ScotRail had bought Twenty-seven IC125 from Great Western Railway when the latter greatly reduced its fleet of the type. The trains got equipped with automatic doors and enclosed toilet systems as well as new interiors to allow continued operation before entering service with the new owner. The first ScotRail IC125 entered service in 2018, with the unit involved in the accident starting service on the 9th of April 2020, 4 months and 3 days before the accident.

The first ScotRail IC125 being presented to the public in 2017.

The accident

The night between the 11th and 12th of August 2020 saw severe thunderstorms hit the area around Aberdeenshire, causing localized flooding in various places and disrupting a number of rail lines, including flooding at Perth Station (where a train was trapped in the water) and the Falkirk line (which flooded when the bank of Union Canal failed).

Flooded streets (left) and rail lines (right) at Perth on the morning of the 12th of August 2020.

At 6:38am Inter7City from Aberdeen to Glasgow departed Aberdeen station right on schedule. Due to the ongoing Covid-19 pandemic, which saw Aberdeen under lockdown at the time, the train carried just nine people. At 6:55am the driver received a message from another train up ahead, reporting that a landslip had blocked the line between Carmont and Laurencekirk. The train was held at Carmont station for two hours while an employee of Network Rail, the owner of the tracks, came out and arranged for the train to cross into the opposite tracks so it could head back to Aberdeen. By 9:36am 1T08 departed Carmont station for Aberdeen, by that time the weather had changed to bright sunshine. After leaving the station it picked up speed, reaching 117.2kph/72.8mph according to the data-logger. This was well below the local speed limit of 121kph/75mph.

At 9:38am, just 2.3km/1.4mi after changing direction, the leading power car ran into muddy soil that obstructed the tracks, lifting the leading locomotive’s leading bogie out of the tracks. Momentum kept the train going straight ahead as the track curved to the right, destroying the parapet of a small bridge as the locomotive started to turn, splitting up the train. A group of workers who had been at the bridge managed to narrowly avoid being struck by the derailing train, making them both the first responders on scene and the first people to notify emergency services.

The leading power car fell down a wooded embankment, turning over 180° to the direction of travel before coming to a rest just as its rear section burst into flames. The impact with the parapet sheared the driver’s cab off the locomotive, killing the driver inside as the cab broke apart. The first passenger car rolled over as it turned 90° sideways, coming to a stop with the second passenger car (ending up inverted) and the fourth car on top of it. Damage to its leading end from running into the leading locomotive compromised the car’s structural rigidity, with the subsequent rollover essentially folding the walls in on the car. This meant a near-total loss of survival space. The Mark 3 passenger cars don’t feature any retention-system for the bogies in a vertical direction, causing them to quickly separate from the derailing train if the vehicles lifted up while the Class 43 locomotives feature a heavy duty wire meant to keep the bogie from separating too easily. The second car remained largely intact as far as the basic structure is concerned, but was pierced by a detached bogie. Car 3 also went down the embankment just past the leading locomotive. The rear locomotive remained on the rails and didn’t fall over, suffering some damage from running into the back of car 4 which it was still coupled to.

Three people died in the derailment, with the other six being injured, three of them severely. The driver had died in the detached cab of the locomotive, the conductor in the leading passenger car as it rolled over and one passenger from the second car was killed when he was thrown from the train as the car rolled over.

Aftermath

The conductor who had been riding as a passenger in the rear passenger car survived the derailment with minor injuries and proceeded to walk 4.5km/2.8mi along the tracks to a trackside telephone to notify the local signal box. In the meantime the workers who had been at the bridge contacted emergency services and used the What3Words-app (a geocode-system meant to identify any location with a 3m/10ft precision) on one of their phones to send responders to the exact location, which was located on a wooded hillside between fields with no proper road nearby. Rescue helicopters from the Scottish Ambulance Service and the Coast Guard were the first responders to reach the site, followed by three Coast Guard vehicles who drove to the site on the tracks from Carmont station.

Cars 1, 2 and 4 photographed after the accident, with smoke from the burning locomotive in the background.

Photos taken at the site at 10:22am showed that the third car had caught fire at some point after the accident, by 11:09, when the fire was finally addressed, it had engulfed a significant part of the car. Luckily, by this point, the car had long been evacuated. The fire was later traced to auxiliary batteries carried on the right hand side of the car’s underpinnings, which had been shaved clealy off the car in the derailment. Each train car carries 48 lead-acid batteries in polymer boxes, remains of which were found scattered among the wreckage. The fire consumed the batteries almost entirely, making it impossible to tell how exactly their fire had started, but later examinations of identical batteries made it likely that physical deformation during the derailment had caused the damaged batteries to heat up, with the lead plates’ thermal mass (the ability to absorb energy) delaying the ignition of the fire. Once the fire started the train car’s final position, resting uphill with the battery box at the lower end, aided the flames as they found their way inside the car through an HVAC-vent and then readily found material to burn as the fire spread uphill. In the end about 2/3 of the car burned down to a shell, with responders managing to keep the fire from spreading to the surrounding brush.

A photo from the report showing the train car on fire at 11:00am.

The fire that started on the leading locomotive was traced to a different cause, with spilling diesel fuel spraying out of ruptured fuel tanks and coming into contact with an undefined piece of hot metal (from the engine or possibly the brakes), which managed to ignite the “foggy” cloud of fuel. The fire destroyed the (already severely damaged) engine compartment and moved right up to the bulkhead behind the driver’s cab, but failed to penetrate through it. Had the cab not detached from the locomotive in the accident this would’ve kept the driver safe from the fire.

The burned wreckage of the leading locomotive as shown in the report.

Investigators managed to recover footage from a surveillance camera in the cab of the leading locomotive, which recorded for long enough to show brownish debris covering the northbound track, be it barely. The footage matches the footage recovered from the camera in the rear locomotive, which recorded throughout the derailment of the train. Calculations based on the footage and the aftermath resulted that the rails were covered by no more than 170mm of debris. This was enough to push the wheels up just enough for the flanges on the inside of the wheels to end up on top of the rails, unable to make the train follow the track as it started to curve. It is assumed that the driver either didn’t see the obstruction at all, or that he overestimated the amount of debris the train could cut through without problems. The latter theory is somewhat backed up by groves in the debris showing that 1T08 managed to cut through most of the material without derailing.

Screen captures from the train’s CCTV-system as shown in the report.

The debris that ended up derailing the train was traced to a nearby “French drain” installed in 2012. A french drain is a small trench filled with gravel or smaller rocks, covering a perforated pipe meant to collect surface water which is then directed elsewhere in a controlled manner. The drain at the site of the accident consisted of a 450mm/18in pipe running along the edge of a field uphill from the tracks for 306m/1003ft, sloping down to track level on the last 53m/174ft. Along its path 19 maintenance-hatches allow access to the pipe, each one creating a small reservoir within the pipe. In 2008 a landslip had blocked both tracks of the rail line, leading to Network Rail planning a large-scale improvement of protective measures along the route in 2009. Engineers precisely calculated every aspect of the drainage system to be installed at the site of the accident, down to demanding the gravel covering the pipe to be 20–40mm in diameter, with no more than 5% of the material (by weight) being smaller than 10mm. However, a post-accident examination found as much as 10% of the material in samples to be too small.

One of the maintenance-hatches photographed after the accident, completely unearthed.

The drain started at a small creek which ran towards the rail line, going parallel to the tracks until eventually ending in an open ditch near the bridge. The heavy rainfall during the night prior to the accident overloaded the drainage system, causing water to run along the gravel ditch above the pipe in a temporary creek (some of which can be seen in the surveillance-footage pictured above). This meant that gravel from the ditch along with random lose soil and stones from the adjacent field was carried along the trench above the pipe, which increasingly washed out as the rain kept falling, focusing the direction and location of the material. Furthermore, the soil and smaller gravel additionally clogged the perforated top of the pipe, increasing the amount of water going through the trench and washing it out. Eventually the washout reached the rail line, spilling over the tracks.

The failed drainage and debris photographed 55 minutes after the derailment. CP18 is a maintenance-hatch.

It is unknown when exactly the drainage-failure became catastrophic, but modeling and calculations resulted in a time-span of 45 minutes between 8:15am and 9:00am being the most likely point when the track got covered in washed out debris. The failure was aided in the hillside at the site of the accident being somewhat funnel-shaped, guiding relatively large amounts of water (and with that soil/gravel) towards the drain. This was known, but it has to be assumed that Network Rail underestimated how bad rainfall at the site could get when they approved the plans for the drain.

An aerial view of the area surrounding the site of the accident.

With the cause of the derailment found in the failure of the drainage-system alongside the track investigators turned to the consequences. The question was if the age of the train and with that the outdated engineering made it unreasonably dangerous for passengers and staff. With Network Rail wanting to repair the line and resume service the train cars and as many of their pieces as the investigators could find were recovered from the site a few days after the accident and taken to a warehouse to be further examined. The recovery necessitated a 900m/980yd road to be built into the forest next to the tracks so 600 metric ton crane could be constructed for recovery of the train cars. The British Army used a tracked armored recovery vehicle to drag the train cars within reach of the crane, which loaded them onto flatbeds to be hauled off to a secure warehouse at Glasgow. A main focus was of the examination would be the aged construction, especially that of the leading locomotive, where the driver’s cab had departed the train as a whole before breaking apart as it continued down the hillside.

A temporary platform was constructed to allow a heavy duty crane to come in and recover the train.

The cab of the Class 43 is largely made of glass fiber reinforced plastic (GRP) and then bolted to the bulkhead ahead of the engine-compartment as well as to the frame of the locomotive. In contrast to most modern trains it features no steel or aluminum skeleton that merely gets covered in plastic for aerodynamics and protection from the elements. The impact of the leading bogie with the bridge’s parapet (which tore the leading bogie from the train) was enough to tear the cab off its bolts, causing it to fly forward and collide with the ground at an estimated 72kph/45mph. With the driver’s seat featuring no restraint-systems it is entirely possible that the driver already suffered fatal injuries at this point, being violently flung forwards.

The cab broke apart on impact with the ground, with the roof, windscreen and left hand wall coming to a rest 22m/72ft from the rest of the locomotive. The floor, both doors, control desk and unoccupied seat were found spread out along the hillside. The report notes that the all-plastic construction of the cab was certainly inferior to more modern constructions with a metal “roll cage” of sorts, but with even the most current crash-standards for British trains only calculating protection in collisions with immovable objects at up to 36kph/22mph a modern train’s drivers cab likely wouldn’t have fared much better. It would likely have remained attached to the locomotive, but a total loss of survival space is possible. Furthermore, with no restraint-system, any somewhat violent collision is bound to send the driver flying into the windshield/control desk. Way back after the Ladbroke Grove accident in 1999 the possibility of adding seatbelts or airbags to high speed trains was the topic of a lengthy investigation of its own, with similar efforts starting again in 2003. However, both times it was estimated that the cost of implementation was grossly disproportionate to the benefits, especially considering the likelihood of train drivers choosing to ignore provided seatbelts.

The leading end of the leading locomotive as shown in the report, with the cab cleanly missing. A photo of the rear locomotive’s cab is provided for comparison.

Next up the investigators looked into the Mark 3 passenger cars making up the train, noting that, as seen in other accidents, they mostly performed relatively well in the accident, limiting the loss of survival space in most of the train even during rollovers and resisting penetration by bogies and debris to a high degree, even at the windows (which commonly shattered, but often remained in their frames regardless). A consequence of the cars’ age was that they had no dedicated crumple-zones designed to safely absorb energy in a straight collision (end to end). The vestibule (the area around the doors, between the seats and the end of the car) is protected by four vertical pillars on each end of the car, two at the corners of the car and two pillars next to the end of the aisle. On the first passenger car all four pillars were sheared off as the leading end of the car ran into the back of the locomotive, causing a total loss of survival space in the vestibule which killed the conductor. Two passengers were ejected from the destroyed train car and survived.

The leading end of the first passenger car after being recovered from the site.

Mark 3 passenger cars were designed to withstand certain lateral forces at various heights above the car’s floor, but beyond the pillars there is no structural rigidity integrated into the design, making the body of the train car easy to deform/compress once the pillars fail. Furthermore, Mark 3 passenger cars don’t feature doors between the vestibule and the seating-area, meaning in case of a train separating during derailment or even ripping open passengers are more likely to be ejected from the train car. This happened during the derailment, with the two passengers ejected from the leading passenger car avoiding being struck by the derailing train after being ejected late into the derailment-sequence, surviving with serious injuries.

Lastly, the investigation found advanced corrosion around the attachment-points of the pillars, credited to the cars’ advanced age. Rust repairs had been carried out on the pillars when the cars were upgraded before entering service with ScotRail, but with no photos of the work and the pillars having been largely destroyed it was impossible to determine the condition of the pillars ahead of the accident. The report also notes that the exact forces the pillars had to cope with couldn’t be calculated, making it impossible to tell how much of a role the corrosion played in the car’s behavior in the accident. The pillars’ mounting points had rusted away to the point of their thickness having been reduced from 5mm to as little as 3mm. In the end the report merely notes that rust can play a significant role in reducing the structural rigidity of the train cars and should be closely monitored going forwards. Wabtec, the company who had refurbished the cars for ScotRail, calculated that the corrosion could take away as much as 20% of the material before a notable decline in structural rigidity would set in. In some spots the material-loss by corrosion was worse, but the report notes that the forces of the derailment were so violent that even a fully intact pillar would have likely separated from the frame.

Severe rust on the remains of the leading car’s leading end as pictured in the report.

The report goes on to say that, due to being made prior to 1994, the Mark 3 cars don’t feature any vertical bogie-retention. All but one of the train cars’ bogies separated from the train, and since nearly all of their mounting-points showed little damage it was concluded that most bogies departed the train when pitching-motions (the train cars leaning left and right) freed them from their pivot-points. During a derailment it is favorable to keep the bogies attached to the train cars, as this faster slows the train cars (with the bogies digging into the ground) and detached bogies commonly becoming obstacles for following train cars, where penetration of the car body can cause severe or lethal injury to occupants. The second passenger car struck one of the detached bogies with its left hand side, severely indenting and tearing the wall and moving interior furnishings. However, due to the near-zero occupation of the train no passengers were injured by the damage, and while one passenger died when being ejected from the car’s vestibule the three survivors aboard the car managed to evacuate the car on their own.

Damage from a bogie (center) and another train car (right) penetrating the second train car.

The investigators created a scenario of a more modern train suffering the same accident, and found several points where advances in engineering MIGHT have reduced the severity of the outcome:

  • Modern train cars have anti-climb elements at the end of each car to prevent train cars mounting each other’s frames in an accident and feature extensive crumple-zones designed to absorb energy without a too rapid loss of survival space. These structures could have greatly reduced the loss of survival space on the leading end of the leading passenger car.
  • The IC125 is fitted with “Alliance”-type couplers which aren’t as rigid as more modern systems, and which did indeed fracture during the derailment. A stronger coupler which either wouldn’t have failed or would’ve failed later would have reduced the range of motion of the individual train cars as well as reduced the scattering of the train cars.
  • The missing bogie-retention made the cars more likely to roll over, made it harder for them to slow down and created additional obstacles for the following train cars.

The report emphasizes that this is no guaranteed scenario but just speculation, with too many factors playing into every individual accident to certainly say which difference would’ve caused which outcome.

Responders standing next to the leading passenger car as it sat on the hillside.

The report was published on the 10th of March 2022 (a week prior to this piece being written) and places blame on several factors. Mainly Carillion (a construction company commonly hired for governmental projects) had failed to construct the drainage-system in accordance to the design, which already was barely sufficient. Secondly, the crew meant to inspect the rail line after the severe rainfall had an excessive workload and could not spend enough time on each section to ensure proper inspection and estimation of the safety. Thirdly, the age of the train, despite having been refurbished just weeks prior, played a significant role in the outcome of the derailment. The accident was one of the worst in recent British History, had the train been remotely full it would have become an unimaginable tragedy.

Smoke billows from the wreckage after the accident as the train burns.

The rail line was rebuilt after the accident with an enlarged drainage-system, with Network Rail upgrading their design-requirements for railway lines in areas that are at an elevated risk of flooding. At the site of the accident so-called gathering rails and guard rails were installed on the tracks, designed to keep a derailed train’s wheels aligned with the tracks. This measure is meant to make it much more unlikely that a derailing train would miss a bridge, hit the parapet or head down the embankment as the track curves.

The site of the accident after repairs, with the new rails intended to increase operational safety.

On the first anniversary of the accident rail services across Scotland stopped at 9:43am for a minute of silence, honoring the victims of the derailment. The same day a plaque was unveiled at Stonehaven station, reminding people of the accident and its victims.

The memorial plaque, mounted on an exterior wall at Stonehaven station.

The leading locomotive and the passenger cars involved in the accident were scrapped once the investigation concluded, with the rear locomotive (#43030), which had only suffered minor damage in the accident, being kept as a parts-donor before it will eventually be scrapped as ScotRail chose to not reuse it, partly due to its involvement in the accident. The remaining IC125 are still in regular service with ScotRail with 4 or 5 Mark 3 passenger cars per train.

The Dundee-Aberdeen rail line is scheduled to be fully electrified by 2035, which might put an end to ScotRail’s use of the diesel powered IC125 on the line.

ScotRail #43030, the rear locomotive from the accident, photographed at Dundee in June 2020.

_______________________________________________________________

Join the discussion about this post on Reddit!

The index of all my posts on Reddit.

--

--

Max S

Train crash reports and analysis, published weekly.