Poor safety practices, management oversight cited as factors in track worker death

The National Transportation Safety Board (NTSB), the independent federal entity charged with investigating significant railway accidents, has released the final report of their investigation into a June 2017 incident involving the fatality of a LIRR track worker in Queens Village on the day of the Belmont Stakes.  On the morning of that Saturday, train 7263 from Huntington Station to New York struck and killed a track foreperson as it was travelling through the area at about 78 mi/hr.  As the train on track 3 (the westbound local track) approached the crew of four workers standing in the gauge of track 1 (the westbound express track), three of the workers stayed put, but the foreperson stepped into the path of the train on track 3 at the last minute and was struck and killed (a fifth worker was the watchperson and was standing clear of the tracks).

The crew was standing in the gauge of track 1 when the
foreperson stepped into the path of the train on track 3 (red
circle) and was fatally struck (Photo: NTSB)

The NTSB determined the probable cause of the of the incident was LIRR's poor track worker protection practices, poor overtime management processes that allowed track employees to work long shifts without properly considering and mitigating the risk of fatigue, and poor management oversight of track worker protection safety practices.

The incident occurred on the morning of the annual running of the Blemont Stakes, when tens of thousands of people flood LIRR trains to Belmont Park.  The incident caused significant delays for thousands of passengers headed to the Belmont Stakes, as the incident investigation tied up QUEENS Interlocking (the critical junction leading to and from the Belmont Park spur) and resulted in 72 late or canceled trains.  The crew was on duty and supposed to be on standby at Queens Village so that they could respond quickly in the event of a track issue that might impact service, but two hours into their shift, LIRR's assistant supervisor of track instructed the foreperson to conduct a walking inspection of the main tracks through Queens Village.  While the crew did stop and conduct a job briefing after their task changed, this task seems to have been an unnecessary distraction on an already busy day.  The NTSB said its its report that deciding to inspect and foul the tracks when it was known that there was going to be increased train traffic due to the Belmont Stakes was "an unsafe management decision".

Poor track worker protection practices

A watchperson holds a yellow whistle marker and alerts
workers and approaching trains (Photo: MTA/Patrick Cashin)

The LIRR's train approach warning system was principally faulted for the incident, with the NTSB saying that it "fails to ensure adequate protection for roadway workers" on tracks where trains are allowed to continue operating.  This practice, where a watchperson alerts road workers of an approaching train, increases the risk of an accident, particularly when there are multiple tracks and limited areas for workers to safely clear up and wait for a train to pass by.  Regulations currently allow for train approach warning systems to be used in cases where workers foul live tracks for incidental inspections and minor repairs.  The NTSB issued safety recommendations to both the FRA (R-20-006) and the MTA (R-20-008) calling on both parties to amend their regulations and rules to identify or define when using this type of warning system creates an unacceptable risk and prohibiting its use during those circumstances.

The train approach warning method of track worker protection is not considered very robust, since it relies on someone standing watch to alert working employees of approaching trains on otherwise active tracks.  Other mechanisms of track worker protection, including setting signals to prevent trains from entering the work area or employing Positive Train Control to stop trains before entering the work zone area are considered safer (one of the specific requirements of PTC is to protect employees and equipment working on the track).  PTC was not in service on this segment of track at the time of the accident.  In this case, no provisions were put in place to prevent trains from entering the work areas

The on-track safety briefing conducted by the work crew was found to be incomplete, with critical information (such as the predetermined place of safety where workers could retreat to as a train approached) not discussed.  The corners cut by the track foreperson, who ultimately lost his life in the incident, also involved failing to recognize that using train approach warning protection to occupy the tracks in this instance created an unacceptable risk.

The LIRR's right of way through Queens Village is elevated, and as a result the side slopes at the outer edges of the local tracks are steep with heavy vegetation at ground level, which would have made it difficult, if not impossible, for workers to clear out of the right of way as a train approached.  In this case, the work crew agreed to clear approaching trains by moving to an active main track, despite that being prohibited by FRA regulations and LIRR rules.  Interviews with the employees involved in the incident found that this apparently was a common practice, happening multiple times in the past, including earlier in the morning on the day of the incident.  The NTSB's investigation also found that violation of track worker protection rules were routine and not consistently recorded by supervisors.

Some of the third Track mitigation efforts will leave scant
room for track workers to clear active tracks (MTA/LIRR)

The importance of having safe places for track workers to clear to is very important when designing track projects.  The Main Line Grade Crossing Elimination project, for example, managed to do away with requiring any residential property takings by squeezing the third track into the existing right-of-way for most of the length and using retaining walls to limit the need for side slopes.  But this limits the amount of shoulder space that track workers have to safely clear the tracks, meaning that more straightforward tasks like inspections and minor repairs will require outages that disrupt service (instead of being done on in service tracks between trains).

Fatigue due to overtime mismanagement

It is believed that the track foreperson who lost his life during the incident and the watchperson that was standing guard were fatigued because their overtime shifts did not allow for adequate periods of restorative sleep in the period leading up to the incident.  The NTSB believes that had the railroad employed practices to better manage and mitigate the risks associated with worker fatigue, requests for overtime shifts on prior days would have been denied and the likely fatigue avoided.

Hours of service statues and regulations (which restrict how long employees in certain roles can remain on duty) do not apply to trackworkers, and while the NTSB recommended the FRA adopt scientifically based hours of service requirements, the ultimate responsibility for mitigating this risk falls to the individual railroad.  Contract terms agreed by and between the LIRR and SMART, the union representing trackworkers, allow employees to schedule overtime based on skill and seniority, without consideration of fatigue, which exposes employees and the public to unnecessary risk.  The NTSB issued recommendations to both the MTA (R-20-009) and SMART (R-20-010) that they work together to develop and implement an improved work scheduling program.

In this case, the foreperson involved in the incident was on duty and working day and night the two days prior to the incident, meaning he had no real opportunity for any restorative sleep the two days before the incident occurred.  In addition to his regularly scheduled 7:30a-3:30p day shifts, this employee also worked additional shifts on overtime from 8:00p Thursday to 7:30a Friday, 11:00p Friday to 7:00a Saturday, and 7:30a to 11:00p on Saturday (leaving less than half an hour of down time before the shift where the incident ultimately took place).  Add these all together, and the foreperson was on duty for 38 out of the 50 hours prior to the incident.  Had this person finished his shift as scheduled, he would have worked a total of 52 hours over the course of three calendar days.

The working hours of the employee involved in the incident (NTSB)

Fatigue can cause workers to take shortcuts, delay their reactions, and make poor decisions.  The NTSB has for many years implored railroads to use scientifically-based principles to reduce the risks of fatigue when assigning work schedules.  As part of its investigation, the FRA ran the employees' working conditions through one of its certified mathematical fatigue models, and the results showed that both the foreperson and the watch were "below acceptable levels of effectiveness on the day before the accident" based on their work schedules.

The findings of this incident yet again establish a link between the LIRR's poor management and oversight of overtime and increased risks to passenger safety.  Track workers are often towards the very top of the list of people in raking in excessive overtime payments.  The employee who tragically lost his life in this incident took in more than $151,000 in overtime payments in 2016, more than doubling his $100k base salary.  In a derailment last Memorial Day weekend just east of Speonk, an investigation by the FRA into the incident revealed that federal inspection reports were likely falsified by a signalperson responsible for visually inspecting the track circuit that eventually failed and resulted in the accident, and that the matter was the subject of a criminal investigation.  In that case, the signalperson (identified as Stuart Conklin, of Ronkonkoma), was already the subject of a criminal probe over allegations that he lied about his working hours.  A source told the New York Post that "'[h]e basically made his own hours,'" and that "his no-work approach was an open secret among his colleagues", all while raking in tens of thousands of dollars in overtime.

As these types of incidents prove, excessive overtime not only hurts our wallets, failing to effectively manage overtime hours also introduces significant, unacceptable risks to the safety of employees, passengers, and the general public.

Management oversight deficient

The NTSB also chided the railroad for poor management of its situational awareness testing program (called SAFER), finding that it was not identifying known areas of non-compliance with LIRR rules and FRA regulations and that the program was ineffective with ensuring compliance with track worker protection rules and regulations. (FRA inspectors missed the compliance failures, too, though that is not a legitimate excuse).  LIRR management was not consistently recording instances where track worker protection rules were violated by employees, and thus the railroad's program was ineffective in ensuring compliance with these important rules and regulations.  The railroad's policies require supervisors to take corrective action in the form of personal instruction upon noticing an instance of noncompliance, and take disciplinary action when there are repeat violations.

LIRR management was also cited for not making a more deliberate assessment of the hazard of having employees occupy the tracks during the period of increases service, nor did railroad management adequately assess weather the train advance warning system used was an acceptable form of protection given the circumstances.

This was also not the first time in the United States where track workers ended up being struck by trains because they cleared into active tracks and did not properly assess the risks associated with this type of protection.  Earlier that year, two BNSF track workers were killed in Edgement, South Dakota as they were standing in the track gauge on active tracks after the train advance warning system in use in this case failed to provide adequate notice to them to clear out of the right-of-way.

Snowballing of unforced errors

While it has thankfully been decades since a LIRR passenger last lost their life in an incident on the railroad, reading these types of incident investigation reports really causes one to shake their head...  There are several concerning common threads that extend across many of these recent incidents, including:

• Multiple points of failure and human error that are small and seemingly insignificant on their own, but snowball and end up compounding risk
• Failure by management to make sure that workers were actually doing their jobs, and properly following all rules and regulations while doing so 
• Sloppy operations with delays and disruptions requiring last-minute changes to assigned tasks, the risks of which are not always properly reviewed and mitigated
• LIRR's insular culture resulting in it not properly learning from safety incidents on other railroads, recognizing similar risky situations on our own system, and proactively implementing measures to prevent incidents before they happen 
• Failure to implement safety protocols and risk mitigation programs in a timely manner, like the installation of Positive Train Control or widespread sleep apnea testing of operating employees

The common thread across many of these recent incidents, including this one, the Speonk sideswipe, and the Brooklyn-Atlantic Terminal derailment earlier in 2017 is a snowballing of unforced errors.  It's not like these are freak accidents, it's a result of cutting corners combined with on-the-ground workers and management not doing their best to proactively mitigate risks before incidents happen.  And this is only what we learn from the events that lead to major incidents that prompt third-party investigations...heaven only knows what's going on day-to-day that we don't hear about...

The LIRR is still, on the whole, very safe, and rail travel remains one of the safest ways to get around...but as we have tended to see in recent years, it doesn't always seem like the railroad is trying its hardest to keep it that way.  I have expressed significant concerns in the past over incident investigation protocols and the importance of recognizing points of failure, implementing measures to correct them, and being vigilant for other areas of weakness.  It may seem silly to harp on little things like incomplete job briefings, not stopping to reassess risks when circumstances change, or certain groups of employees that work excessive amounts of hours for the sake of raking in OT...but as we have seen several times now, it's the little stuff that can really screw things up.