2016 Croydon tram derailment
On 9 November 2016, a tram operated by Tramlink derailed and overturned on a sharp bend approaching a junction. Of a total sixty-nine passengers, there were seven fatalities and sixty-two injured, nineteen of whom sustained serious injuries. This was the first tram incident in the United Kingdom in which passengers died since 1959.
The service was running from New Addington to via Croydon, and was on the approach to Sandilands tram stop soon after 6:00. The second interim report into the accident indicated that although the speed limit approaching the junction was, the tram had been travelling at approximately. Driver error was found to be the cause of the accident.
Accident
The accident took place in the dark and during heavy rain at 06:07, on a sharp left curve approaching the points where the route from New Addington converges with the line from Beckenham Junction and Elmers End. The curve is located in a cutting, which comes almost immediately after the line emerges from a series of three tunnels on a straight section of track after leaving Lloyd Park tram stop. It has been described as a "sharp bend", and has a speed restriction.The tram entered the curve at a speed of approximately and derailed, overturning on its right side and coming to a stand beyond the point of derailment, damaging the side of the tram and ejecting several passengers through broken windows. The tram involved was No. 2551, a Bombardier CR4000 constructed in 1998 by Bombardier Transportation, a two-section articulated unit with a maximum design speed of.
The emergency services confirmed that 51 people were injured, and initially stated that five had died. The death toll later rose to seven. The victims, six men and a woman, were between 19 and 63 years old. Sixteen of the survivors were described as having serious or life-threatening injuries. The final casualty figures were seven dead and 62 injured, 19 seriously. One person was uninjured.
It was the deadliest tram accident in the United Kingdom since an accident at Dover in 1917 caused eleven dead and 60 injured, and the first in the United Kingdom in which passengers died since three people died in an accident in Glasgow in 1959. It is also the deadliest accident on any rail network in the UK since the 2001 Great Heck rail crash.
The UK news channels and media was criticised for not reporting the crash as they were reporting wall to wall coverage of the 2016 United States presidential election.
Aftermath
The injured were taken to St George's Hospital in Tooting, and to Croydon University Hospital. As well as the 51 people taken to hospital, seven more made their own way to hospital for treatment. Twenty-two ambulances, eight fire engines, and over 70 firefighters from the London Fire Brigade were sent to the scene. Assistance was given by Croydon Council, the Red Cross, Salvation Army and railway chaplains.FirstGroup, which operates the tram service on behalf of Transport for London, said it was "shocked and saddened by what happened." An extra minute was added to the two-minute Armistice Day silence at Croydon Cenotaph. TfL later made an offer to cover the funeral expenses of the victims.
During the night of 10–11 November, the tram was righted in preparation for removal from the accident site, and it was removed on the morning of 12 November. The tram had suffered serious damage to its right side, onto which it had overturned.
After the accident, no services operated on the line between East Croydon and Addington Village, Harrington Road or. Partial tram services ran between East Croydon and, between and Addington Village, and between and Harrington Road. Full services were reinstated on 18 November.
It was found that the speed restriction sign for the bend at Sandilands was not visible to drivers until the tram had travelled for past the point where braking would need to have been initiated to reach the required speed at the sign — drivers "were expected to know this from their knowledge of the route". At Sandilands, an additional speed restriction was imposed before the curve, and chevron signs were installed to give better warning of the curve. Chevron signs were also installed at three other locations on the Croydon Tramlink system. Following recommendations made in the first interim report into the accident, tram systems in the United Kingdom introduced stepped speed reductions where there was a required reduction in speed of or more. This affected systems in Blackpool, Edinburgh, the West Midlands and Nottingham.
In March 2017, it was reported that Tramtrack Croydon Ltd and TfL had admitted liability for the accident. Victims and survivors would not have to sue for compensation for losses caused by the accident. In August 2017 TfL confirmed they would be altering Tramlink timetables to take account of reduced speed limits.
In November 2017, two days of strikes by tram drivers were announced following the installation of fatigue monitoring devices. The devices shine an infrared light into the driver's face and are capable of generating an alert and vibrating the driver's chair if eye movements indicate a lack of attention. Some drivers have raised concerns about health and safety issues, and have described the device as a "spy in the cab".
On 22 January 2018, the Office of Rail and Road organised a safety summit in Manchester to discuss the findings of the Rail Accident Investigation Branch report on the investigation into the accident. In January 2019, it was announced that all trams on the system were to be fitted with a system which would automatically apply the brakes if the tram exceeded the speed limit.
As of July 2019, tram 2551 was still in the custody of the RAIB in Farnborough and had not returned to service.
Investigations
British Transport Police
The 42-year-old tram driver was arrested by the British Transport Police on suspicion of manslaughter. After questioning, he was released on bail until May 2017. According to Metro, one aspect of the police investigation was whether or not the tram driver fell asleep; and The Guardian reported that some passengers said the driver had blacked out at the controls. In November 2017, it was reported that a file was being prepared for the Crown Prosecution Service, who would decide whether or not a trial would take place. it was reported that the CPS were still awaiting the file from the British Transport Police, who were still gathering evidence. On 31 October 2019, British Transport Police and the CPS announced that neither the driver, operating company or TfL would face prosecution. The CPS stated that while there was evidence of negligence by the driver, it did not count as "gross" and so "manslaughter by gross negligence" could not apply. Prosecutors also stated that because the section of tramway where the crash happened was neither legally a railway or a public place then various other potential offences would not apply.A former driver suggested, to The Times, that a blackout was a possibility, due to the erratic shift patterns that the drivers had to adhere to. He also said that the vending machine at the tram depot was stocked only with energy drinks, and that "Nobody is ever fully awake; I was always in a bit of a daze and that is because the way the shifts work doesn't allow the drivers to get a regular sleep pattern." Following this, multiple sources reported on a video apparently showing a different driver struggling to stay awake at the controls. The driver concerned was suspended pending an investigation into the matter.
Victoria Derbyshire programme
In April 2017, the BBC Two programme Victoria Derbyshire reported that its own investigation into drivers falling asleep at the controls of trams on the Tramlink network revealed four such cases. Six drivers claimed that the dead-man's vigilance device fitted to the trams was not fit for purpose. Tramlink stated that the devices were "fully functional".Office of Rail and Road
The ORR opened its own investigation into the accident, concentrating on whether or not safety rules were followed. They confirmed that British trams are not fitted with an overspeed protection system. The ORR was expected to make an announcement about its investigation early in 2018. On the third anniversary of the accident, the ORR tweeted that its investigation was ongoing.Rail Accident Investigation Branch
The RAIB also investigated the accident, with data from the tram's on-board event recorder being analysed. The RAIB stated that initial indications suggested that the tram was travelling at a significantly higher speed than permitted. Rail magazine reported that the tram's electro-magnetic track brakes had not been activated. Following the accident, The Guardian reported that on 31 October passengers had made allegations on Facebook of a tram travelling round the curve at excessive speed. The Evening Standard reported an earlier passenger complaint describing the tram as "tipping" on the curve. The driver of the tram involved in the incident of 31 October was not the one involved in the accident on 9 November.The interim report was released a week later on 16 November 2016. At the time of the accident it was dark and it was raining heavily. There was no evidence of any track defects, or obstructions on the track, that could have contributed to the derailment. Initial investigation did not indicate any malfunction of the tram's braking system. Initial findings were that the tram was travelling at approximately at the time of the accident—far exceeding the speed restriction. The RAIB interim report noted that "a tram approaching the Sandilands Junction area from Lloyd Park at would need to brake at its full service rate of 1.3 m/s² for approximately 180 metres before the speed restriction board in order to be travelling at 20 km/h when the board was reached." The On Tram Data Recorder indicated that some braking had occurred within this distance but only sufficient to reduce the tram's speed from 80 kmph to 70 kmph.
A recommendation was made that a further speed limit should be introduced prior to the one for the curve at Sandilands Junction before the line reopened to traffic. The recommendation was accepted and three further speed restrictions were put in place before the line reopened. The report also found that a number of passengers with fatal or serious injuries had been ejected from the tram through broken bodyside and door windows. In November 2016, Rail magazine called for the RAIB to complete its investigation and release the final report "much more rapidly than has become the norm".
A second interim report was released on 20 February 2017. In April 2017, it was reported that there had been three cases of speeding on the section of line which included the accident site in the period November 2016–April 2017. In one case, a tram was reported to be travelling at in a zone. Within the second interim report the RAIB noted that drivers of trams approaching the curve could be expected to sight the curve and the speed restriction sign from with full beam headlights and with dipped beam headlights. Although the report states that the tram's braking system was not capable of slowing the tram sufficiently between the point where the speed restriction sign became visible and the point at which the speed restriction is enforced, it also says, "There was no sign to indicate to drivers where they should begin to apply the brake for the Sandilands curve; they were expected to know this from their knowledge of the route."
In the light of the first interim report the UK government Office for Rail and Road requested that all operators of light rail tramway systems apply a system of stepped speed restrictions where reductions in speed greater than are required by changes in the characteristics of the track.
The final report was published on 7 December 2017. Driver error was found to be the cause of the accident. The most likely scenario being that the driver had a microsleep episode approaching the bend. Fifteen recommendations were made. Key findings of the RAIB investigation were that the tram's windows, which were made of toughened glass, were not strong enough to contain passengers inside the tram. All the fatalities had been ejected through the right side windows. Laminated glass would not have broken, improving the survivability of the accident. Regulations applying to trams were similar to those applying to buses. If railway regulations had applied, laminated glass would have been fitted. The tram's operators had not considered it possible for a tramcar to overturn. There had been insufficient reporting of previous incidents by drivers, who feared that they would be disciplined rather than such reports being seen as an opportunity to learn a safety lesson. The system for dealing with complaints from passengers was not fit for purpose. The investigation also found that trams were not as safe as the ORR previously thought, having a higher accident rate than other rail transport and also buses and coaches.
An update to the final report was issued on 24 October 2018. The update included the addition of a Transport for London audit of Tram Operations Limited's fatigue risk management system, a copy of a letter sent to all UK tram operators and some minor technical corrections. In response to Recommendation 1 of the final report, the Light Rail Safety and Standards Board was established. The ORR has observer status on the LRSSB.