Sleepless Staff: The 2016 Hoboken (USA) Derailment
Hoboken is a city of 60416 people (as of 2020) in the northeastern USA, located in the federal state of New Jersey 11.5km/7mi east of Newark and only 1.5km/0.9mi from New York City, which lies across the Hudson River (both measurements in linear distance).
Hoboken Terminal is an 18-track 9-platform dead-end train station located in the eastern part of Hoboken, right on the shore of the Hudson River. Since opening in 1907 the station was a major transportation hub from the start, combining train, ferry, subway and streetcar-lines with buses later joining the hub too. Nowadays the station is one of North America’s busiest train stations, seeing over fifty thousand people per day with 16 thousand of them being rail passengers. Most services at the station are operated by NJ Transit, the state’s government-run public transportation provider.
The train involved
PVL train number 1614 was a commuter train service from Spring Valley (New York) to Hoboken. On the day of the accident it consisted of a so-called cab-car, a train car with driver controls at one end, two regular passenger cars and a locomotive pushing from the back. The three train cars were Alstom Comet V, with cab car number 6036 leading the train. The type was introduced into service with NJ Transit and Metro North (a main rail service provider in New York) in 2002, with 265 units being made. Each Comet V measures 25.9m/85ft in length at an empty weight of 45.4 metric tons, with the regular cars seating up to 117 passengers in a single-class configuration. The cab-cars feature a small driver’s cab for remote control of the locomotive at one end, reducing seating capacity to 109. Push-pull services with a cab-car leading the train on part of its journey are far from new or unusual, allowing far more efficient operations than having to move the locomotive to the other end of the train before the train can start its return-trip. At the time of the accident Train #1614 carried approximately 250 passengers and a crew of 3.
Pushing the train at the time of the accident was NJ Transit locomotive number 4214, an EMD GP40PH-2B. The type was introduced in 1993 by converting former freight-only GP-40 locomotives for passenger usage (such as adding the ability to provide power to the train cars). By the time of the conversion the locomotives had been in service for as much as 28 years. The GP-40 is a four-axle diesel locomotive measuring 16.76m/55ft in length at a weight of 113 metric tons. They can reach a top speed of 105kph/65mph.
On the 29th of September 2016 PVL train #1614 is departing Spring Valley station (New York) cab-car first at 7:23am local time, running under the command of 48 years old Mister Gallagher. There is no automatic train control on the line, it’s down to the train driver to watch signals and other instructions in order to maintain the required speed. The train’s journey was largely uneventful, making fequent stops at various stations along the line. However, as it entered Hoboken Terminal at 8:38am people at the station quickly realized that something was wrong. The train was coming into the dead-end track 5 too fast, and wasn’t slowing down. Moments after entering the station cab car #6036 crashed through the rigid buffer stop at the end of the track, ripping the steel structure out of the ground as it mounted the platform at 34kph/21mph.
The train proceeded across the platform, crashing through a wall and crossing the station concourse, only stopping when it ran into the wall separating the concourse from the station’s waiting area. Damage to the walls and roof support columns caused part of the station roof to collapse, including a large steel beam which fell through the cab car’s roof. 110 people, mostly aboard the train, are listed as injured in the report, while a woman waiting at the platform was fatally struck by falling debris and died at the site.
The cab car actually stayed largely intact as it went through the buffer stop, through the first and into the second wall, but pieces of the falling roof caused severe damage, tearing open most of the car’s roof. The station was completely shut down right after the accident, with fears that more of the roof might come down around the collapsed section. Mister Gallagher survived the accident with undefined injuries and was interviewed by investigators shortly after the accident. He didn’t recall anything being wrong with the train, but his memory skipped from when he approached the station to when he regained consciousness on the floor of the leading car’s cab after the accident had taken place. Investigators couldn’t find any sign of a technical malfunction when examining the train at the site. They recovered the first of two data-loggers the day after the accident, whose data ended up being unusable from physical damage, so they had to wait until early October to recover the backup data-logger after the train had been removed from the station.
Data recovered from the second data-logger showed that the train had been travelling at 13kph/8mph just a minute before the accident, upholding the 16kph/10mph speed limit. But 22 second later the driver accelerated to 34kph/21mph, despite the speed limit still being in effect. An emergency stop was finally triggered just one second before the train went off the rails. As such Mister Gallagher was clearly the one at fault for the accident, the question that remained was why he had acted the way he did.
Mister Gallagher had passed his most recent medical examination in July of the same year, but investigators found that the records were incomplete. During the examination the doctor had noted symptoms of sleep apnea, a disorder which causes the patient to pause breathing intermittently while asleep, reducing blood oxygen and thus disrupting the sleep-phase’s restorative capability. This can lead to chronic sleepiness, focus issues and other related problems due to simply not getting enough relaxation out of the sleeping hours. If it persists over longer periods of time it can also lead to “microsleep”, where patients randomly fall asleep for seconds at a time. NJ Transit has a system in place to refer drivers with symptoms for the disorder to a specialized sleep-study, but there was no evidence of Gallagher ever being recommended to seek out such measures.
Five days after the accident Gallagher was submitted to a full physical examination and sleep study. Firstly, his weight, missing from previous examination’s documentation, was recorded, finding him morbidly obese at 146kg/322lbs. A suspicion that the weight was left off the paperwork to make it look better could not be proven. Furthermore, he was diagnosed with a severe obstructive sleep apnea (OSA), a form of the condition where excess tissue drops back and blocks the patient’s airways rather than a neurological cause.
Going over the train’s data-logger and talking to passengers and train staff again revealed several smaller errors made ahead of the accident, such as skipping the mandatory horn-signal ahead of level crossings or sounding the horn late. There were no radio-conversations and no evidence of Gallagher using his phone while on duty, leaving these errors as indications of insufficient focus caused by the effects of undiagnosed OSA. The final report suggests that by the time the train pulled into the station Gallagher’s focus was so poor that he moved the throttle lever in the wrong direction by accident, accelerating rather than slowing the train.
Following the discoveries about Gallagher’s medical status the investigation also reviewed the remaining train crew’s medical records, finding that both the conductor and assistant conductor also fulfilled the requirements to be referred for suspicion of sleep apnea, but had also never been referred. At the time of the accident it was the individual decision of the doctor performing the examination wether or not an employee in a safety-sensitive position (such as train crew) would be referred or not. As such the report concludes that, while the accident was the direct fault of Mister Gallagher, NJ Transit’s failure to follow a recommendation by the Tri-State Medical Society Task Force played a significant role in the accident, as he may not have ended up in the condition which caused him to cause the accident if he had been referred, diagnosed and treated after his last pre-accident physical examination.
With the cause of the accident determined as human error another question moved into the focus of the investigation. Why was there no automatic train control system that would stop the train if the driver failed to do so? PTC (Positive Train Control) had been in widespread use in the USA for several years, but was nowhere to be found at Hoboken Terminal. The Rail Safety Improvement Act of 2008 (RSIA) had required all operators of intercity or commuter rail services to get a PTC-system online by December 2015. That deadline was later extended to 2018 and could be extended two more years if the rail service providers can prove that they reached certain milestones in the preparation, such as installing all the hardware required. However, the investigators found that there are at least 35 passenger train stations in the USA which end at platforms the way Hoboken Terminal does, and all 35 station’s operators had successfully requested an exception from the introduction-deadline, meaning they were not equipped with PTC and it wasn’t planned to equip them. NJ Transit in particular had declared Hoboken Terminal’s 20 dead-end tracks as “other than main line track”, successfully arguing that the permanent-red signal and low speed limit were sufficient in ensuring trains would stop ahead of the end of the tracks.
With the PTC-exemption in place Hoboken Terminal’s tracks don’t fulfil the “no single point of failure”-principle most modern railroads operate by, since a single point of failure (in this case the driver) can cause a catastrophic accident. After the accident NJ Transit defended their choice, explaining that the measures they had in place were generally sufficient. In consequence of the accident the speed limit for incoming trains was cut in half to 8kph/5mph and it was dictated that there had to be at least one more member of the train crew riding in the cab as the trains pull into the dead-end tracks of the station. This, in NJ Transit’s claim, removes the possibility of driver error causing the trains to overrun the buffer stop.
As a last point to note the investigation pointed out that Hoboken station was still using the original steel-and-concrete buffer stops from when the station had opened in 1907 rather than a more modern design. A representative of the buffer stop’s manufacturer explained to investigators that the design in place was perfectly sufficient under the assumption of a low-speed impact of a train that either rolled freely or was braking. However, train #1614 was under full power as it went through the buffer stop, simply ripping the structure out of the ground and taking it along for the rest of its ride. The report notes that a more modern design of buffer stop with hydraulic dampening or a slide-mechanism (where the whole buffer stop can move a short distance to slow the train hitting it) could make it less likely for trains to overrun the buffer stop while also not stopping a train so suddenly that there would be a needless risk of injury to passengers. A downside of more modern systems is that they require more space, reducing usable platform length.
The report closes with several safety-recommendations to both the Federal Railroad Administration (FRA) and NJ Transit. Among them is the recommendation to the FRA to introduce stricter rules regarding the operation of passenger services on dead-end tracks and one to NJ Transit to shift their medical referral system from the doctor’s personal decision to a centralized committee to reduce the influence of personal sympathy. The report also reiterates the recommendation to the FRA to make examinations regarding sleep apnea and similar disorders mandatory for safety-sensitive railway workers.
History repeats itself
The investigation into the accident at Hoboken Terminal had barely started when a near-identical accident happened at the Atlantic Terminal in Brooklyn, New York City. In the morning of the 4th of January 2017 a commuter train overran the buffer stop at the station, mounted the platform and ran into the adjacent wall, leaving 108 people injured. The accident was caused by a fatigued driver being unable to sufficiently focus on his tasks, with the fatigue being traced to severe undiagnosed sleep apnea. The accident was included in the report on the 2016 Hoboken Derailment, with the investigation extending the same recommendations to the Long Island Rail Road, the provider involved in the 2017 accident, as they did to New Jersey Transit, as the cause was exactly the same.
A kind reader has started 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. Please give it some love, I’m very thankful for the posts being published there.