Wednesday, December 06, 2017

Guest Blog: NTSB missed the mark

Written by  Chip Kraft, for Railway Age
Guest Blog: NTSB missed the mark

The NTSB analysis of the 2016 Amtrak backhoe accident on the Northeast Corridor at Chester, Pa., that killed two m/w workers and injured 39 has missed the mark. This incident is by no means the first of its kind.

NTSB issued Safety Recommendation R-08-06, after a 2007 accident near Woburn, Mass., in which two Massachusetts Bay Transportation Authority (MBTA) maintenance-of-way employees died: “Require redundant signal protection, such as shunting, for maintenance-of-way work crews who depend on the train dispatcher to provide signal protection.”

In 2013, NTSB reiterated Safety Recommendation R-08-06 to FRA after an accident in which a Metro-North maintenance-of-way employee was struck and killed by a train in Connecticut.

And now, as a result of the 2016 accident—the third accident in less than a decade under remarkably similar circumstances—NTSB reiterated the same recommendation a third time. Stronger action is needed to prevent future incidents, particularly in Positive Train Control (PTC) territory, or else the 2016 accident most likely will not be the last.

As part of R-08-6, the mention of shunting as part of the recommendation is, unfortunately, a serious distraction. Shunting is not considered failsafe by the railroad industry, and the practice can cause other problems. From an FRA report on roadway worker protection (downloadable at the link below):

“In response to FRA’s request for comment regarding a potential redundant protection requirement, AAR, NTSB, SEPTA, BMWED/BRS, APTA, MTA, NJT, and an individual, submitted comments. NTSB urged FRA to add a provision in this final rule requiring using redundant forms of protection such as shunting. AAR urged FRA not to adopt such a provision, indicating it would be counterproductive from a safety perspective. AAR stated such a provision would be counterproductive because shunting cannot be relied on due to: (1) the characteristics of track infrastructure that lead to periodic loss of shunt for certain equipment; (2) the susceptibility of shunts to work only intermittently when used near signal islands; and (3) the lack of reliability of individual locomotives or roadway maintenance machines to shunt. AAR’s comment pointed to the safety issues shunting presents in some circumstances, specifically grade crossing warning device malfunctions and signal system interference, and to concerns related to cost, training, and the practicality of shunting requirements (e.g., trying to shunt as a roadway worker conducts walking track inspections or mobile weed spray operations).”

Nor is applying GPS devices onto maintenance vehicles, as suggested by recommendation #2 of the NTSB report (downloadable from the link below), a fully effective solution. David Schanoes addressed this issue directly:

“So ... getting back to Chester, Pa., we can in fact put GPS receivers on track equipment. As a matter of fact, I bet lots of m/w equipment is already equipped with GPS receivers so the movements can be tracked for budgetary purposes. But even with that, it’s a timing issue. Does the dispatcher see the occupancy indication in time?

“We can even integrate the occupancy indication with the train control system, so that the obstruction of the track triggers a downgrade of signal or authorized speed indication, but again, that’s a timing issue: the train is rounding the curve at 110 mph, a hi-rail vehicle gets on the track 300 feet on the other side of the curve, it takes one second for the obstruction to be processed and transmitted to the train, two seconds for the train brakes to set up, (even without any ‘free run’ time), and guess what? Location/occupancy technology to the contrary not withstanding, the phone’s ringing and it’s not someone yanking your chain.”

Rather than simply reiterating itself, NTSB should have paid more attention to the AAR’s objection and should have revised recommendation R-08-6 to remove the mention of track shunting. Instead, a more effective solution that would actually solve the problem should have been recommended. The right approach to “Require redundant signal protection (R-08-6)” would be to make Employee in Charge (EIC) or TFT (Track Forces) terminals mandatory for all PTC installations.

By using EIC terminals to request and obtain electronic work authorities, engineering employees would have a ready means to ascertain whether they have the track or not, and the PTC system would protect the work crews, to ensure that appropriate routing restrictions and speed limits are enforced. Once a track force has the EIC or TFT, then the electronic authority could be relayed electronically to the GPS on board each of the work crew’s vehicles. This would allow an equipment operator to be warned if they are approaching the limit of or exceeding their authorization. Conceivably, the foreman could also be notified.

BNSF’s Hy Rail Compliance system already integrates GPS data with track occupancy authority data and could be a model for the industry on how to effectively integrate GPS data for protecting work forces.

Currently, FRA does not require EIC terminals on the Northeast Corridor, since Amtrak only uses Exclusive track authorities, not Permissive authorities that trains can pass through. FRA has only been requiring EIC only for Permissive work authorities in Vital PTC territory.

Currently, FRA does not require EIC terminals on most freight railroads that are using I-ETMS, since I-ETMS is only approved as a “non-vital” PTC system. FRA has not been requiring EIC terminals as a condition of non-vital PTC certification, since the requirement for such certification is only to improve safety, not to provide an absolute assurance of safety.

As a result, only a handful of PTC railroads are actually using EIC terminals today. NTSB should have recommended that FRA make EIC terminals mandatory for all PTC implementations. If this is not possible, then FRA should at least make EIC terminals an absolute requirement of Vital PTC certification without any exceptions. In conjunction with EIC terminals, railroads could and should additionally implement supplemental protection to the level of the individual workers, as SEPTA and Metro North have already done on their respective territories.

Chip Kraft is Director of Operations Planning at Transportation Economics & Management Systems Inc.

Get the latest rail news

Rail news and analysis from Railway Age, IRJ and RT&S by email