Ten90 Solutions LLC

While we're waiting for the NTSB to figure out a way to access the event recorder on the NJT cab car before the data is lost or corrupted,  let's jump ahead a year or two and anticipate the conclusion of NTSB's investigation and the list of forthcoming recommendations.

(As an aside: even without the event recorder data, we can determine the speed of the train.  Hoboken terminal is an interlocking, with track circuits of fixed length; data should be available showing the elapsed time taken between the train occupying one track circuit and the next successive track circuit. Simple math will then provide an average speed of the train over the route, and railroading is a simple business.)

It's November 2017, and the NTSB is conducting an open hearing, live streamed on the web, to consider the findings of its investigation into the overspeed derailment of NJT Train 1614 on September 29, 2016.  

The investigation has revealed that the train speed when entering train shed track 5, and throughout its operation prior to the emergency brake application, was consistently between 17 and 20 mph; that the throttle position varied between 0 (idle) and +1 (minimum power); that the cab signal properly displayed "restricting;" that the cab signal audible indicator had been properly acknowledged when the cab signal had first changed to "restricting;" that the headlights were properly displayed; that the train struck the bumping block at a speed of 19 mph; that the impact with the bumping block caused the locomotive engineer to move the brake handle to the "emergency" position, that the train traveled 106 feet after the emergency application was initiated.  

The NTSB agrees with the assertion of its vice-chairperson that the primary cause for this overspeed derailment is the failure of NJT Rail Operations to install PTC in Hoboken Terminal.

The NTSB finds that contributing causes to this derailement are:  (a) FRA's approval of Main Line Track Exemptions for passenger terminals where speeds are restricted to no more than 20 mph and interlocking rules are in effect (b) failure of the locomotive engineer to properly control the speed of the train due to unknown reasons that may include, but are not limited to, obstructive sleep apnea, fatigue, pre-existing medical conditions, transient ischemic attack, distraction, distraction due to background radio communications between other trains and the control center, etc. etc.  

(Note: side bets accepted and covered on how close I am to the actual data values recovered from the event recorder, if recovered).  

NJT Rail Operations "failure" to install PTC in Hoboken Terminal is no failure at all.  It is in accordance with the fundamentals of railroad regulation-- develop and adhere to the minimum requirements.

Basically, it's pointless to find fault with a railroad for meeting the statutory requirements for safe train operations, which NJT Rail Operations have met in the past, continue to meet in the present (leaks about FRA audits to the contrary notwithstanding) and will meet in the future. 

Which gets us to the MTEA exclusion from PTC requirements for passenger terminals. The basis for FRA's incorporation of passenger terminals within the MTEA process is the expense, effort, technical difficulty that railroads will encounter in kitting out terminals where highly complicated, and compressed, switch and signal arrangements govern the movement of trains.

"Interference" among the numerous electronic transmitters used to enforce limitations on train movements makes installation of full PTC protection at or near these switch and signal configurations impractical .  

The application of the MTEA to passenger terminals also expresses FRA's "ambivalence" to restricted speed.  For certain aspects of certain regulations, FRA seems to accept "restricted speed operations" as a functional equivalent for positive protection.  For example, in 236.1005 Requirements for Positive Train Control,  FRA allows "A PTC system shall enforce restricted speed over any switch where...wayside or cab signal system or other similar appliance, method, device or system of equivalent safety, requires to train to be operated at restricted speed."

This, according to FRA satisfies the provision in RSIA 2008 requiring PTC to prevent movement over an improperly lined switch, which is to say, the law and the regulation are different.  The regulation says the law is satisfied if the train is allowed to operate through the improperly lined switch at less than 20 miles per hour. The law says no such thing. 

Similarly, FRA allows "a PTC system shall be considered to be configured to prevent train-to-train collisons within the meaning of paragraph (a) of this section if trains are required to be operated at restricted speed and if the onboard PTC equipment enforces the upper limits of the railroad's restricted speed rule (15 or 20 miles per hour)."  

Translation: the law requires PTC to prevent train to train collisions.  The regulation does not; the regulation requires only that such collisions be limited to 20 mph for any train so involved.

Strangely enough or not, FRA has determined that restricted speed operation does not satisfy the requirements for PTC systems to provide an "appropriate warning or enforcement when: a mandatory directive is issued associated with highway rail crossing warning system malfunction"-- known in the business as "stop and warn."  

In these cases, according to those who have participated in discussions with FRA on the matter, enforcing restricted speed instead of a positive stop at the grade crossing is unacceptable.  Full stop must be enforced.

I don't know how many fatalities have occurred at grade crossings due to locomotive engineers operating at restricted speed rather than coming to a full stop, waiting for an employee to establish flag protection at the crossing and then proceeding, but there's plenty of evidence that train to train collisions at 20 mph do produce fatalities.  In any case, I'll check FRA's safety database and get back to you, probably before NTSB effects the download of the event recorder from train 1614.  

The issue might be one of consistency to those with closed minds, but my mind is open.  The issue is that the automatic enforcement of the speed to no more than 20 mph has absolutely nothing to do with the essential requirements of the rule governing operating at restricted speed. 

The essential, determining characteristic for operating under a signal indicating "restricting" is that the locomotive engineer must be prepared to stop within one-half the range of vision--short of a train, obstruction, improperly lined switch, or broken rail.  The numerical value, whether 15 or 20 or even 10 is simply designed to limit the damage if the locomotive engineer fails to comply with the operating requirements.  

So here we have FRA sanctioning the failure to operate a train properly as operating the train properly.    

To be sure, there's a lot to be said for capping the speed of a train when operating under the restricting requirement.  The less speed, the less energy. A train moving at 10 mph embodies 1/4 of the energy embodied in a train moving 20 mph.  The  fact is that 20 mph is far too great a speed for passenger train operation in a terminal of such complexity and density of switch and signal equipment that installing PTC is either too expensive or technically difficult or both.  

The very conditions that present an obstacle to PTC installation are the conditions that make enforcing a numerical value for restricted speed at 20 mph inadequate to the task of safe train operations in the terminal.  

So what is to be done?  First and foremost, because it is expensive, complicated, or otherwise burdensome to fit out the entire interlocking, to equip every track, every signal with full PTC, does not mean it is burdensome to either (a) enforce a reduction of speed to 10 mph throughout the entire terminal without the other functions of the PTC system or (b) enforce a reduction to no more than 10 mph on platform tracks or (c) enforce a zero-mph target at the extreme ends of all platform tracks.  All of these things can be accomplished through the proper location of a series of transponders in the case of ACSES systems, or wireless interface units for railroads using the I-ETMS platform. 

I estimate that if train 1614 had been restricted to 10 mph it would have traveled only 40 feet, transmitting much less energy to the structures located on or near the platforms.  

How do I know?  Simple math?  Yes,  And the hard way.  When I was superintendent in Grand Central we did have a locomotive engineer who operated an AM peak inbound train at 10 mph, straight into a bumping block.  

The bumping block was demolished. The train did not hit the passenger platform "apron" area beyond the bumping block.  No one was injured. The train traveled 39 feet from the time the emergency application was initiated. 

And BTW, we got the download off the event recorder 11 minutes after the derailment occurred

Right now MTEA exemptions have been approved for Washington DC Union Terminal, Grand Central Terminal, Penn Station New York, Jamaica Interlocking complex in Queens, Boston South Street Station, New Orleans Union Passenger Terminal, Chicago Union Station, Springfield Terminal, Harrisburg, Pa and others.

We all want to make sure this never happens again.  Choose (a) (b) or (c), or any combination thereof.  Or make your own version, but do not expect enforcement of 20 mph operations to provide the necessary level of safety.  

David Schanoes

October 2, 2016, 

Nothing is worse than having an itch you can't scratch.  Leon Kowalski, Blade Runner