T-Broken: The 2007 Kerang (Australia) Level Crossing Collision

Max S
14 min readOct 29, 2023

--

Background

Kerang is a town of 3893 people (as of 2016) in southeast Australia, located in the federal state of Victoria 123km/76.5mi north of Bendigo and 50km/31mi southeast of Swan Hill (both measurements in linear distance).

The location of Kerang in southeast Australia.

The town lies on the Piangil railway line, a 225km/140mi non-electrified single tracked rail line running from Yungera to Bendigo. The line opened in 1926 to carry both passenger and freight services and is constructed in Australian broad gauge, meaning the rails are 1600mm/5.3ft apart rather than the usual 1435mm/4.7ft.

The site of the accident seen from above today. The train approached from the northwest (top-left of the image) while the truck was coming from the south (bottom of the image).

The Vehicles Involved

Train number 8042 was a so-called “V/Line” (the slash is part of the name) passenger service from Swan Hill to Melbourne, consisting of three four-axle passenger cars pulled by V/Line N460. The N-class is a six-axle diesel locomotive introduced in 1985 in both regular and broad gauge versions. Each N-class locomotive measures 18.87m/61.9ft in length at a weight of 123 metric tons and can reach 130kph/81mph. The locomotives all carry the name of a city along with their number, with N460 being christened “City of Castlemaine”.

V/Line N460, the locomotive involved in the accident, photographed a few months after the accident.

The three-car train pulled by N460 consisted of 3 VicRail Type N passenger cars, a family of four-axle passenger cars introduced in 1981. The leading car behind the locomotive was a ACN21 first class car, followed by a BRN20 second class car with a buffet section and a BN19 second class car without a buffet section. Each train car measured 22.86m/75ft in length at an empty weight of 44 metric tons. The train had a capacity of 221 passengers, but only carried 34 at the time of the accident (6 in the leading car, 21 in the buffet car and 7 in the rear car) along with two conductors and a driver. The relatively high capacity stems from the broad gauge allowing rather wide cars, leading to a 3+2-seating arrangement in second class (called “economy” by V/Line).

A Type N passenger car identical with those involved in the accident, photographed at an unknown date. The pictured unit is a second class car (like BN19 was).

On the day of the accident a Mister Scholl (aged 49) was scheduled to drive a semi truck from Victoria to the neighboring state of South Australia using Murray Valley Highway. The truck he used was a 1999 Kenworth K104 Aerodyne, a three-axle “cab-over” (meaning there is no long hood in front of the driver’s cab) truck weighing 8.9 metric tons with 150l/39.6gal on board. The truck was pulling a Krueger refrigerated curtain trailer measuring 13.57m/44.5ft in length at an empty weight of 9.2 metric tons. The trailer is fitted with a refrigeration unit at the outside of the leading wall and, at the time of the accident, carried 14.5 metric tons of medium density fiberboard (wooden boards made of compressed wood fibers) along with 2 more tons of what the report refers to as “miscellaneous freight”.

A K104 Aerodyne similar to the one involved in the accident (left) and a sketch from the report showing the trailer with some measurements in millimeters (right).

The accident

Mister Scholl passes northbound through Kerang on the 5th of June 2007 at around 1:30pm, performing a regular transport between Wangratta (Victoria) and Adelaide (South Australia). On the day he was making his first trip after a four week vacation break. By 1:30pm he was driving on the Murray Valley Highway, a two-lane paved road running through a flat, open landscape. The highway crosses the Piangil railway line 5.9km/3.7mi down the road from Kerang at an angle of approximately 40°. At the time of the accident the crossing featured signage, bells and red warning lights, but no barriers.

The V/Line train from Swan Hill to Melbourne triggers the level crossing’s lights and bells at 1:33:46pm, approaching from the northwest at 91kph/56.5mph. Mister Scholl fails to react accordingly, only hitting the brakes when he’s about to cross the tracks. The truck does slow down, but the momentum of the heavy vehicle and cargo means it doesn’t lose speed fast enough. The locomotive and leading car make it through the crossing unharmed before, at 1:34:11pm, Mister Scholl’s truck runs into the broadside of the train’s second car at an angle of 32°, still travelling at 65kph/40.4mph.

The collision pushes the back of the second car off the tracks, tearing open its side wall along with the forward right hand side wall of the third car. The coupler between the rail cars fractures in the collision, separating the last car from the train as it derails. The truck suffers extensive damage in the collision, partially separating the tilt-cabin (the driver’s cab can be tilted forwards to allow maintenance access to the engine) from the frame. The trailer pushes the stopped truck aside before running into the train as well, breaking up under the forces of the impact. Mister Scholl survives the collision despite the extensive destruction dealt to the cab of his truck. He suffers severe injuries, as do 22 people on the train. However, 11 passengers are killed as the train is torn open, ripping out several seats.

A photo from the report giving an overview of the aftermath.

Aftermath

The first call to emergency services connected 8 seconds after the collision, another 5 calls followed in the next three minutes including one from the train control center who had been alerted by the train driver. The first responders arrived at the site by 1:50pm. Five people were airlifted to Melbourne while the remaining injured survivors were taken to Kerang by ambulance.

The truck as encountered by first responders, with the cab lying “face down” in front of its frame.

It was obvious from the start of the rescue effort that the consequences of the collision could have been far worse. The train was very lightly occupied at the time of the accident, with the most crowded car (BRN20, the second car) carrying 21 passengers, less than a third of its capacity. Tragically, BRN20 was also the car which was struck by the truck, suffering the worst damage by losing over half its right-hand wall along with several rows of seats being shoved back through the car or being torn from the train entirely. Ten of the eleven victims had been seated in BRN20, nine of which on the right hand side where the truck hit the train.

The interior of car 2 after the recovery of victims finished. Note that a few seats that had been torn loose were removed to access victims, but the open side is a consequence of the accident rather than responder access.

Side impacts are a known weakness of passenger train cars, with no real practical solution. A lengthwise impact is cushioned by the buffers at the end of the train car and by energy-absorbing elements in the vestibules (door-area, where no passengers sit), aided by equal-height frames in a lot of rail cars which allows the sturdy frames to transfer energy that would otherwise crumple the body of the car. However, there just isn’t that much space for side impact protection. Modern passenger rail car bodies are often built similar to a ribcage, consisting of vertical struts to which a relatively thin metal skin is attached, with insulation, wiring and interior dressing taking up the remaining wall thickness.

The truck managed to penetrate BRN20’s bodyshell by means of its trailer’s frame cutting through the sheet metal skin, tearing an increasingly large gash into the car as momentum kept pushing the trailer into the still-moving train. The truck itself had broken up on impact, absorbing forces as it was being pushed aside by the trailer. The trailer’s frame, in contrast, was rigid, had no crumple zone and sat slightly higher, allowing it to mount the frame of the rail car.

Photos from the report showing the initial rupture of BRN20’s right hand wall (left) and the nearly roof-high opening the rupture escalated to (right).

Rigidity of rail cars against side impacts could theoretically be improved with (in simplified terms) added bracing across the interior of the car and/or thicker material for the walls. However, this would drastically reduce interior space and also remove the ability to move along the train without having to step out of the train car. Furthermore, a completely rigid body design would leave its occupants at the mercy of high, potentially fatal acceleration forces on impact, as there is nothing to absorb energy. It’s the same reason why modern cars feature a crumple zone designed to compress, rather than being built to maximum rigidity. The more energy a vehicle can absorb in a collision without compromising survival space, the less severe injured its occupants are likely to be. Also, one should keep in mind how rare blunt “T-Bone” side-impact accidents are on rail lines. Level crossing accidents almost always have the train plow into the side of another vehicle, and there are extremely few places in the world where railway tracks intersect at a blunt angle.

Side-impact collisions between trains usually occur at a sharp angle on a set of points, and while train cars tend to lean or even fall over during them they rarely lead to the bodyshell being torn open.

The aftermath of the 2014 Train Collision at Mannheim (Germany). Two passenger cars fell over, but their impact-side walls remained intact.

The investigation into the accident at Kerang quickly figured out that the accident was the fault of the truck rather than the train, zeroing in on Mister Scholl’s actions. This was made a little more difficult as Mister Scholl decided to not talk to the investigators, leaving them to rely on limited information from his statement to the police. Skidmarks at the site showed that the truck had had its brakes fully applied approximately 49m/161ft ahead of the crossing. The marks were caused by the wheels of the trailer, which, in contrast to the truck, was not fitted with an Anti-lock braking system which would automatically release and re-apply the brakes to keep the vehicle from sliding. The skidmarks veered off the road to the left at approximately 8°, explaining how the remains of the truck ended up on the left side of the road.

The wreckage of the truck sitting at the site, with the detached and derailed rear train car in the background beyond the level crossing.

The severity of the damage being much higher on the truck’s right hand side than on its left led investigators to conclude that Mister Scholl likely realized he wasn’t going to stop in time and tried to drive into the brushland to the left of the road (where the rail line left more space than on the right, due to the angle of the crossing). His truck thus proceeded to strike the broadside of the second train car with its forward right hand corner, crushing a significant portion of the cab before being deflected from the train, making way for the trailer to collide with and subsequently penetrate the body of the car. This impact-scenario is matched by scuffing and broken windows on the forward right hand side of BRN20, assumed to be caused by the cab being deflected off the train car before the trailer managed to penetrate the wall of it further down its side. The impact derailed BRN 20 and brought it out of alignment with car 3 (BN19) as the coupler failed, allowing a partial frontal impact between the trailer and the third train car. BN19 suffered extreme damage , caving in its forward right hand corner and losing most of the passenger survival space within its leading 5m/16.4ft before a 2.5m/8ft gash marks the path taken by the trailer as it departed the train car. This theory was backed up by the compressor from the trailer’s refrigeration unit being found in a footwell partway down BN19.

BN19 sitting in the wreckage with it’s leading wall caved in from the collision with the trailer.

The investigation ended up creating a computer animation of the assumed incident sequence to fill in some of the gaps. The sequence showed that the slight right hand curve of the highway as it approaches the crossing may have played a role in the accident, obscuring the warning lights for a few moments at a distance of 360m/1181ft to the crossing due to the positioning of the windshield frame of the truck while the train is concealed behind a small group of trees. However, the report notes that the warning lights could be brought back into view by the driver moving his head slightly forward.

A screenshot from the investigation’s animation (with notes added by me), showing the truck’s windshield frame briefly hiding the warning lights. Note the yellow warning signs for the upcoming level crossing.

The report notes that the warning lights become visible again as the truck enters the curve ahead of the crossing, being fully visible approximately 9 seconds before reaching the rail line. The truck would have then navigated the curve (maintaining view of the active warning lights) before reaching the straight section to the crossing at a distance of 106m/348ft from the crossing. At this point both the train, traffic waiting on the other side of the crossing and the warning lights were fully visible. At that point the truck was travelling at approximately 100kph/62mph, meaning it covered 27.8m/91ft every second. During the events that led to the accident Mister Scholl thus started braking 2 seconds after leaving the curve. This is within the 2–2.5 seconds usually assumed to be the reaction time for motorists. This means that, due to only starting to brake after leaving the curve, Mister Scholl had no chance to stop in time as the reaction time saw the truck cover vital stopping distance before starting to slow down.

The animation of the truck’s perspective at a distance of 106m/348ft from the crossing.

At a basic level, one might take that knowledge and conclude that Mister Scholl was distracted and simply started braking too late. He couldn’t be speeding as his truck was electronically limited to 100kph/62mph, so he just failed to start braking in time by not paying attention to the warning signs and the level crossing in the distance. Well, as the report proceeds to explain it, things aren’t quite so simple.

Employer records stated that Mister Scholl had been driving for 3 hours when the accident occured, having departed at 10:30am after showing up at 9am. This placed him well within the five hour interval commercial drivers in the state of Victoria had to adhere to when it came to taking breaks from driving. Thus, fatigue likely wasn’t a factor. Neither was distraction by cell phone usage, of which there was no evidence to be found. The investigation recreated the moments ahead of the accident on the 14th of June 2007 and noted that, ahead of the curve, the sun was directly opposed to the truck, causing glare from sunlight reflecting off the road. Furthermore, this meant that the right hand side of the train lay in shade as it approached the crossing, making it even more difficult to spot from a distance.

A photo taken during the reconstruction of the moments ahead of the accident, standing by the warning signs ahead of the curve. Note the significant sun glare along with trees partially concealing the rail line on the left.

The reenactment meant that it can be assumed that Mister Scholl couldn’t see the train or the warning lights until after he passed through the curve, by which point it was too late to slow down. He would have had to decelerate for the level crossing ahead of the curve, on the chance that there would be a train coming. And this was the next puzzle piece. Mister Scholl usually just drove through an empty level crossing. But he was also usually on time.

The trip Mister Scholl was conducting on the day of the accident regularly departed at 10:00am and reached the crossing at 1:00pm. But on the day of the accident an undefined delay during the loading process caused him to run 30 minutes late from the moment he started driving. It’s unknown if he ever saw the level crossing with the lights turned on, but he certainly didn’t expect it. And the routine of the crossing being empty made it an unrealistic expectation that he would slow down because there might be a train for once. He actually said to the police that he spotted the flashing lights AFTER being puzzled as to why cars on the other side of the rail line were stopping.

The truck after the cab was turned upright for recovery. Seeing the damage on the right hand side (where the driver sat) it can be said that he was lucky to survive.

The investigation concludes that he may have suffered a phenomenon referred to as “Inattentional Blindness”, which, according to truck drivers interviewed by the investigation, was far from unknown among long-distance truck drivers. Inattentional Blindness describes the brain filtering what the eyes see, “hiding” things just because we’re not focussed on them. The human brain can only keep track of so many objects in its field of vision, and in a high demand situation (such as driving) the brain might automatically filter out things it deems unimportant. This process is strongly influenced by routine experiences and resulting expectations. The same phenomenon can also show up during extremely tiring tasks, with prolonged boredom leading to a lack of attention. Both of these situations are commonly experienced by long distance truck drivers, and several of the interviewees outright told the investigators that they don’t think they’re fully aware of their surroundings when out on the road.

A pile of seats lies next to the wreckage as responders removed them to access victims.

The report concludes that it is very likely that Mister Scholl fell victim to inattentional blindness, failing to spot and/or properly react to the level crossing warning signs and only noticing the flashing lights when he was already too close to the level crossing for a safe stop. His last ditch effort of trying to direct the truck into the field alongside the road may have worked for a semi-truck on its own, or with an empty trailer, but with 16.5 tons of cargo adding momentum he just couldn’t pull a turn as sharp as it would’ve needed to be to miss the train. The train driver had no chance to influence the accident at all, he likely never even saw the truck before his train derailed.

BRN20, the car with the worst damage, photographed in storage in October 2008.

Mister Scholl appeared in court on the 19th of December 2007, being charged with eleven cases of causing death by culpable (negligent) driving and eight cases of causing injury by negligent driving. The main trial began in May 2009 and ended with Mister Scholl being found not guilty on all charges. However, a number of victims relatives and survivors announced they would pursue civil action against him, with one of the conductors announcing in 2011 that him and Mister Scholl had settled out of court by Mister Scholl agreeing to pay an undisclosed amount.

The level crossing at Kerang was upgraded with moving barriers after the accident, and the warning signs ahead of the curve now feature flashing lights to increase attention. An inquest into this and several more fatal level crossing accidents also led to the design of level crossings and their signage being improved, along with first aid kits and tools to remove seats after an accident being provided aboard passenger trains. However, several victims’ relatives from the Kerang accident have stated that the trial and following safety inquest still left them unsatisfied, as they struggle to accept that, officially, there is no person to blame for what happened.

The site of the accident captured on Google StreetView in June 2023. Coincidentally, glare from the opposing sun can be seen in the captured images.

_______________________________________________________________

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.

--

--

Max S

Train crash reports and analysis, published weekly.