Passenger train derailment in the Bronx

[Nathanael]

And the lazy government railroads like Caltrain and Metro-North?

Well, yeah. And I have to be fair: I just found out that BNSF, alone among the freight railroads, says that it *WILL* comply with the law. The others plan to break the law.

The other railroads who say they are going to comply with the law are all government railroads:

(1) Amtrak, for not only the NEC, Keystone, and Springfield, but also Empire to Schenectady and Michigan corridor

(2) Metrolink in LA

(3) Coaster in San Diego

(4) SEPTA in Pennsylvania

Meanwhile, several of the others, including Metra and the South Shore Line, have *outright stated* that their PTC implementations are being delayed by the class I freight railroads.

Metro-North does deserve particular opprobrium, because implementing ACSES and Pennsy cab signals is not complicated and so there's no excuse for the delay. And LIRR, from what I can tell, just delegated its implementation to Metro-North. (Uh... right.)

All the freight railroads chased the chimera of radio, and were being stupid, which has delayed them, but they're stuck with it now -- should've gone with trackborne data transmission and short-distance balises.

 

[RRUserious]

http://www.cnn.com/2013/12/03/us/new-york-train-crash/index.html?hpt=hp_t1

Now it is union representative making statements about nodding off....

Yes, indeed, it sounds like "Highway Hypnosis" and yes I have experienced it and it is scary like hell.

The really scary thing is that most people don't realize it's happening to them until it's been going on for a long time. People therefore overestimate their abilities.
I've been told that in order to reliably avoid highway hypnosis one should never, ever drive for more than 2 hours -- as an extreme limit -- at a time, without stopping and doing something else. Think about it...

I don't buy that. I've driven across North Dakota and Montana. But never alone. Not sure I'd bother alone. When I have one or more passengers, the social interaction keeps one conscious. Also, when my wife and I did it, we brought along recordings of Balkan songs we both knew. I think this hypnosis must be for solo drivers.

 

[CHamilton]

25 Train Tragedies Could Have Been Prevented in the Past Decade by This Technology

An automated system that could have averted a December 1 commuter rail crash in New York City is stalled in a tug-of-war between Congress and the rail industry

From Scientific American! PTC is suddenly getting a lot of mainstream coverage.

 

[Anderson]

The concern I've heard about PTC is that there's really no ability to override a messed-up signal, among other related issues.

As to the MNRR crash...if he was doing 83, he was speeding for anywhere south of Cold Spring. I have to seriously ask, even ignoring PTC, why there wasn't an automated overspeed stop in place for a train going well outside the speed limit?

 

[jis]

The concern I've heard about PTC is that there's really no ability to override a messed-up signal, among other related issues.

As to the MNRR crash...if he was doing 83, he was speeding for anywhere south of Cold Spring. I have to seriously ask, even ignoring PTC, why there wasn't an automated overspeed stop in place for a train going well outside the speed limit?

MNRR equipment overspeed control is set to max speed of 90mph since they could potentially run at 90mph somewhere in their journey. So there is no means to enforce other civil speed limits.
As for the 30mph on the Spuyten Duyvil curve, signal speed enforcement can be used to enforce that, by jiggering the signal algorithm at CP 10 and CP 12 to display (NORAC) Approach (or the MNRR equivalent) as the most permissive signal entering the block that has the curve in it. Indeed this is what might happen for immediate mitigation. Similar use of signal speed enforcement is used for example on the Back Bay and Elizabeth curves on Amtrak NEC.

 

[RRUserious]

This reminds me of discussions about Air France 447. Particularly the idea of feeding black box info to the ground continuously. The more expert people said data was offloaded at the end of flights for such things as detecting unscheduled maintenance required. Well since this commuter train had a black box, it could be somewhat ameliorating if black box data was systematically collected. The notion is that bad habits by certain personnel could be detected on "safe trips" and used to give additional training where safety standards are not met.

 

[Just-Thinking-51]

When driving my tractor trailer the federal government requires us to stop for 30min ever 8 hours. So the government thinks you can drive up to 7.5 hours between breaks. I have done in the past the full 10 hours (back than it was the limited) with out stopping.

It all about keeping yourself focus on the task at hand.

 

[greatcats]

I too was a bit surprised that the union made comments the way they were stated. As a former railroad employee, I feel terrible for Mr. Rockefeller. I have also been a commercial bus driver and had things go wrong, too. Granted, this is a very serious accident, but apparently he has had a conscientious work record. Life for him must be hell right now.

 

[greatcats]

The highway hypnosis point also recalls the 2011 Amtrak wreck in Nevada that was hit by the truck. One difference with the trucker was that guy had a rather poor driving record. May he too rest in peace.

 

[RRUserious]

When driving my tractor trailer the federal government requires us to stop for 30min ever 8 hours. So the government thinks you can drive up to 7.5 hours between breaks. I have done in the past the full 10 hours (back than it was the limited) with out stopping.

It all about keeping yourself focus on the task at hand.

Of course, you could stop for truck stop coffee. That always helps with that last part. Oh, and then there's CB ("Breaker breaker")

 

[SubwayNut]

Anderson, read the New York Times Article, it provides a really good explanation of how MNR's cab signaling works and how the accident might have been prevented in pull mode where a loud buzzer would have sounded to wake the engineer up from his trance and not push-mode where the only 'dead-man's break' is an easy to engage foot petal. I've seen a printed copy of today's Times and the graphic in the paper is huge!

JIS nailed the immediate solution.

 

[CHamilton]

Lack of Sleep: What It Does to Your Brain

Behind the controls of the Metro-North train that derailed in New York earlier this week was a tired driver, according to new reports that engineer William Rockefeller fell asleep at the wheel.


Could lack of sleep cause such a fatal mistake?...

Early morning hours, like when the Metro-North train crashed, are some of the most vulnerable times for sleepy accidents, Howell said, especially for people whose circadian rhythms favor a later sleeping schedule and make it biologically difficult to function well after waking up with an alarm clock at 5 a.m.

Reports that Rockefeller had been driving for 20 minutes since his last stop and felt zoned out before the accident suggest that he probably fell asleep before the crash, Howell added.

 

[WhoozOn1st]

JIS nailed the immediate solution.

It appears the actual immediate solution is to put two people in the cab:

Federal Agency Orders Changes At Metro-North After Derailment -- http://www.nytimes.com/2013/12/07/nyregion/metro-north-is-ordered-to-modify-its-signal-system.html?hp&_r=0

"Five days after a fatal Metro-North Railroad derailment in the Bronx, the Federal Railroad Administration issued an emergency order on Friday requiring the railroad to have two people in place to operate trains at potentially precarious sites.

"The directive to the Metropolitan Transportation Authority, which operates Metro-North, said the personnel requirement would remain until the railroad updated its signal system to better enforce speed limits."

THEN other changes:

"The order calls on Metro-North to provide the railroad administration with a list of main track locations where maximum allowable speeds drop by more than 20 miles per hour. The railroad is to identify 'appropriate modifications to its existing automatic train control system or other signal systems to enable adequate advance warning of and adherence to such speed restrictions,' the administration said."

 

[Blackwolf]

Sounds like there will be a few new job opportunities opening up in New York very quickly...

 

[ALC Rail Writer]

No alerter in the cab car, or in half the trains operating push-pull:

Doomed Metro-North Train Had Warning System, Just Not in Operator’s Cab -- http://www.nytimes.com/2013/12/05/nyregion/as-metro-north-resumes-service-riders-get-back-to-routine.html?pagewanted=1&hp

"The Metro-North Railroad train that derailed on Sunday included a system designed to warn an operator of a potential accident. But such an “alerter,” which can automatically apply the brakes if an operator is unresponsive, was not in the cab where William Rockefeller apparently fell into an early-morning daze while operating the Manhattan-bound train. It was at the other end of the train." [...]

So while an alerter might have helped prevent the derailment, perhaps not?

An alerter would be the logical countermeasure towards falling asleep, I would imagine that fixing this will be amongst the NTSB's recommendations to MN... Still people fall asleep even with alerters.

 

[jis]

FRA has required MNRR to identify all places on their railroad where speed limit drops from one section to next by more than some threshold (20 or 30mph I forget which), and then modify the signals to use signal speed enforcement to enforce the lower speed. This will naturally take some time to complete. In the interim they are requiring that trains that pass through such areas be staffed with two operators at least through such sections.

This is exactly what I mentioned as the most likely immediate mitigation. In 4 or so years when MNRR gets around to install ACSES II, all this will become a non-issue since ACSES can enforce civil speed limits independent of the signaling

And the good Governor who is as clueless as they come, and would serve himself well if he stopped talking for a bit, has asked LIRR to be forced to do the same without realizing that LIRR has been doing so for the last decade and a half, and using a 7 aspect signaling system to boot, unlike MNRR's 4 aspect one. Politicians! Heaven help us!

BTW, if FRA actually enforces this order seriously, expect MNRR service frequency to be halved, since that is the only way they will be able to double the staff in the cab, at least until they are able to hire and train adequate number of new Engineers. If it applies to Amtrak running on MNRR too then Amtrak might face the same problem on MNRR, so expect to see adverse effect on Empire Service too. They could mitigate it somewhat by allowing a conductor trained person to be the second person, but still there will be adverse impact on train frequencies is what I am hearing.

 

[SubwayNut]

I finally got around to adding a blog post to my website about a walk I took up to Inwood Hill Park to view the derailment. You could see it far and clear across the river and the park was full of genuine press folks with huge telephoto lenses getting there shots under the Henry Hudson Bridge. It took me awhile to decide what I wanted to do with my photos but since I've also gone to witness Metro-North's other 3 major mishaps this year it only seems fitting to cover everything.

 

[Anderson]

I'm just wondering...how much of the "second operator" rule is going to result in a second operator boarding at one station, riding in/out a stop, and then getting off a la the Indiana labor rules back in the 60s? Looking at an MNRR map:

-On the Hudson Line, I see no drops of more than 20 MPH until Croton-Harmon. There's a major drop immediately south of the station (75 to 40). From there, you've got a slow curve into Yonkers (drop from 75 to 50), the infamous Spuyten Duyvil curve (75 to 30, then back to 60 before Marble Hill), and then the drop on your way into Grand Central.

-On the Harlem Line, you have a curve south of White Plains (which could simply be avoided by moving the speed limit increase right north of White Plains to just south of the curve), a possible case just north of Valhalla (a 65/50 drop followed immediately by a 40 curve) and one just south of Mount Kisco (75/45).

-On the New Haven Line, you have a possible case at New Rochelle (one track drops 70 to 30, but I think almost all trains make that stop), a 75/45 at the state line just north of Port Chester, 75/50 just south of Stamford (possibly avoided if you split the drop and move a 75/60 south of the 60 curve right there). You have a 70/45 just north of South Norwalk, 75/40 at the Saugatuck River, 70/45/30 just south of Bridgeport, 45/75 north of Bridgeport, and a 75/30 on the way into New Haven.

--There's a tight 60/40/20 at the start of the New Canaan Branch.

--There's a 75/10/30 situation at the start of the Danbury Branch, but the short distance between South Norwalk and that might have an effect here.

--There's a 75/10/59 at the start of the Waterbury Branch. Further up, there's a 59/25/50 around Derby Junction, a 50/25/50 around Ansonia, a 59/30/59 around Seymour, and the 59/50/30/10 on approach to Waterbury might qualify as well.

It sounds to me like you're going to have a lot of employees doing glorified deadheading from Yonkers or 125th Street into GCT.

 

[MattW]

I'm not sure the 2nd employee requirement is quite as dire as is being said. Most of what I'm going say I draw from lurking in various other parts of the internet. The requirement isn't for a full-time engineer the whole route, just another qualified person in the head end. MNRR already requires a second person when operating at least into GCT from 125th street, not sure if they require one outbound. Getting a 2nd person into the cab when a P32 or Brookville is leading might be trickier though. Another solution the various people floated was for MNRR to step down the speeds so instead of say 70mph to 30mph, it might be 70 to 50 to 30, or so it's not on the razor edge of the requirement, 70 to 55 to 35 to 30. I of course can't speak as to the legality or practicality of these solutions, but it at least seems logical to me.

 

[jis]

Frankly, it should not take them more than a month to address the half a dozen or so locations that are the worst by changing the most permissive signal indication entering the section. When the infamous Back Bay incident happened, AFAIR (the details may be a little off) Amtrak had the signals changed within weeks. FRA had imposed a Stop and proceed at Restricting Speed PSR for that period from the signal at Ruggles all the way to Back Bay (i.e. that automatic signal was always red!). Played enough of a havoc with schedules which motivated Amtrak to fix it post haste.

That trick won't work at Spuyten Duyvil since the signal at CP 12 is a home signal and is not passable like a automatic block signal is.

THE FRA order to MNRR as it goes into effect, will affect all Amtrak trains operating on MNRR territory on the NEC and Empire Service too.

 

[VentureForth]

Looks like the already have a lot of fixes in place: The Daily News

Wonder if these changes will work both from the loco AND from the cab car...

Note the Sperry Doodlebug on the left track. Same company that failed to find a major defect that could have taken out the Silver Meteor a couple weeks back South of Savannah...

 

[jis]

Of course it will work from both ends. How would it not? I am confused. They did exactly what I said they'd do, i.e. change the most permissive aspect of the signal leading into the curve block to Approach (from Clear), which automatically enforces 30mph.

MNRR has only 4 signal speeds it can enforce using its current signaling system, and those are 20, 30, 45 and 90. Anything else has to be manually operated upon, hence the radio verification etc. Boy there will be a lot more chatter on radio until they get PTC in place in 4 or 5 years according to their current plans.

You are assuming that the defect was present when the Sperry unit inspected the track. That is not known so the innuendo is probably uncalled for.

 

[Ryan]

Obviously it's their fault for not being able to see into the future.

 

[VentureForth]

You are assuming that the defect was present when the Sperry unit inspected the track. That is not known so the innuendo is probably uncalled for.

It's not so much innuendo as it is speculation. I'm in no position of authority and isn't that what we do here? Speculate and armchair quarterback?

I would suspect that it is Sperry's responsibility to detect something imminent - not count on a passenger train travelling at 89mph to detect.

Cold weather breakage doesn't happen at a random spot on the track. Fatigue or stress fractures don't begin and manifest in a week. I would suspect that is precisely why a company such as Sperry is hired.

 

[ALC Rail Writer]

Of course it will work from both ends. How would it not? I am confused.

To be fair, until this accident, I thought all the cab cars on pax trains I ever rode on had alerters in them... So it is a valid question if you do not understand the systems in question. Not all of us have your insight into railroad technology