amtrak-csx-trains-collide-in-south-carolina-injuries-reported

[draskouasshat]

I don't agree that PTC contributed . Signal systems are shut down at times for reasons of maintenance or failure at times for various reasons . It is the nature of things that work , that sometimes they do not . Things that are important need backups . Not some well intentioned tower operator saying everything is good to go only because he didn't have knowledge it was not . If the signaling was down could that also mean loco / train detection was down . An accident could have left a car wreck on the tracks and the short ckt it may cause was undetectable . The NTSB failed to mandate a backup nation wide .

Richie, you are completely and whole heartedly wrong. Signal systems do not get shut down from time to time for failure or maintenance. I'm not trying to be an a$$ but this is what i do for a living. Signal systems either work,  or trains don't run. It's set in stone. If somethings wrong, trains move at restricted speed.

For all of you that say ptc wouldn't have stopped this, you have absolutely no idea how complex ptc is or how it actually works in regards to the signal system and you most likely never will.

As for the temporary light, there's so much more to it than seeing up a warning light. We have rules in place when things go wrong for a reason. Rules that are meant to be followed.

[up1950s]

Then we will agree to disagree , and I will not belabor my points .

[dandopinski]

I agree with Mr Draskouasshat. The rules for proceeding in CTC and the rules for proceeding in dark territory are different and well defined. The Employee timetable will tell you that the subdivision you are operating on may have one or the other, or portions of the subdivision that may have CTC on a portion or may have dark territory (think of a dark subdivision that has CTC interlockings or "diamonds" on it). In CTC the only way to pass an Absolute Signal displaying stop is (in Canada anyway) is with written permission and then at restricted speed (not exceeding 15mph and prepared to stop short of switches not properly lined etc..) until a more favorable indication is accepted. In dark territory the Dispatcher issues the Crew written instructions that are repeated back as to how far the Train may proceed and any restrictions en route eg: "be prepared to stop at switch not properly lined at ... meet Train xxx at ...etc

So is this portion of track signalized or is it dark? 

[Maletrain]

Careful there. Mistaking CSX for Conrail in this company could be dangerous...

Oooops!  Corrected.

[pedro]

Maybe 8-10 years ago there was a severe ice storm on the BNSF Chillicothe sub (former Santa Fe west of Chicago) that took out code lines and shut things down completely. They suspended CTC and tried to operate trains by track warrant. While it was a complete clusterf*ck, there is precedent for that at least being possible. I’m a BNSF engineer (on the “BN side”) and got to hear the folly over there first-hand when that happened. The problem is, you’ve got a group of employees who never once in their career ever copied a track warrant. And dispatchers who have never given one. It was not pretty.

The story I’m hearing on this wreck (via sources inside Amtrak) is that’s basically what was going on here. They’re cutting over new signals and running trains by track warrant. Now WHY they’re doing that, I couldn’t tell you. We went through this several years ago on the BN east end (hello @draskouasshat!) and never missed a beat. Well, maybe a little, but we sure weren’t copying track warrants.

Incidentally, we’re up to about 20 Metra (BN) trains a day (out of about 120) running PTC. It takes time. It’s a good system. Not perfect, but it will eliminate many human-factor screwups.

[packers#1]

Adam, please correct me if my "old practices" knowledge is out of date here. From what I'm reading so far, this is a "perfect storm" of errors and inattention:

• Crew from CSX train apparently left switch diverging. Whether this was properly communicated to the DS would be a matter of reviewing radio recordings.
• Signal reported "down for maintenance", which is completely different from "not operating". If taken out of operation it is possible heads were turned, which is to be read as not governing movement. If simply not operating and therefore dark, it would then invoke stop-and-proceed-at-restricted-speed per rules.
• The siding switch would normally throw a shunt across the rails if the siding track is not separately signaled and the switch is thrown diverging, resulting in a "stop" indication for the block. That it is alleged it was thrown and locked indicates that it was a non-signaled siding.
• The shunt should have caused an "approach" to be displayed at the previous signal...
• ...provided that the reason for "down for maintenance" did not also affect signal system communications. It is possible that the "down" could be problems in the cabinet and not with the signal itself, and the block (and shunt) were not being detected to pass the occupancy down the line.
• DS (evidently) gave instruction to the Amtrak head-end regarding maintenance issues at this signal and to expect it to be offline (turned heads), which is different than "ignore the inoperable signal," where rules would dictate "stop-and-proceed, etc."

There's too much we don't know and won't know until the NTSB review. My stomach sunk when I read of this accident, as unfortunately it is shades of a similar accident we had on the SP around 1970. A maintainer "temporarily" jumpered a relay in the cabinet, resulting in a "clear" on the main when the siding switch was diverging into an occupied siding. Three or four crewmen were killed in that collision.

Honestly, this is the type of thing we study in my Engineering classes as an example of how systems go wrong and how human error can defeat even the best systems. In this case, it looks like a bunch of communication that SHOULD have happened that didn't actually occur, and reminds me somewhat of the New York City Blackout of 1977 (https://en.wikipedia.org/wiki/New_York_City_blackout_of_1977), which we studied in class. The root cause in that blackout was an overly complex control panel which didn't show accurate information and a lack of communication where the operator did not ask the right questions nor did others tell him the right information, which lead to him not diverting power effectively to balance the load caused by the lightening strike.

What happened here sounds very similar, with a signaling system offline for maintenance/upgrade/PTC installation, a train operating system in place that no one is familiar with and does not know the proper communication, a switch which was somehow missed by the CSX crew OR a dispatcher lost track of, and incorrect information given to the Amtrak crew, who may have not followed proper procedure for a dark signal (or were instructed to ignore it because the maintenance issue was known and supposedly had been accounted for). I would also assume, given how close the CSX engine was to the switch and the curve to the track, that even at restricted speed, unless the signal was literally right there at the siding, there still would have been a collision had the crew done the stop and proceed routine, since the dispatcher clearly had no knowledge of the misaligned switch.

Of course, we don't know the particulars in this situation, nor what actually transpired, but just looking in with the information at hand, it seems like a lot of communication that was supposed to take place either didn't or did not communicate the proper information in a way that all could understand. Sadly, no matter the complexity or redundancy of the system, these sorts of things happen and all we can really do in the aftermath is learn, build better redundancies and rules, and continue to improve in the hopes that the next accident is due to something else going wrong and not the improved system we put in place.

[draskouasshat]

Do we know each other @pedro ?lol

As for what youve heard pedro, if they went to track warrant while they were doing a signal cutover on a freaking single track sub, those guys might want to hire new exempt employees!

Drasko

[draskouasshat]

Then we will agree to disagree , and I will not belabor my points .

Richie, its like saying gravity makes water fall up. You can't just agree to disagree on that. You don't know enough about signal systems, the ptc system, operating rules, etc to make that decision. Ptc isn't even a gps based system like the media outlets are making it out to be. Its signal based sand depends on the signal system plain and simple.
Honestly the media and their experts should STFU about ptc because they don't know what the f they're talking.

[wcfn100]

Ptc isn't even a gps based system like the media outlets are making it out to be.

There are three PTC systems that use GPS as a base.   

Jason

[draskouasshat]

There are three PTC systems that use GPS as a base.   

Jason

They use gps as an input, but inputs from the signal system, if it has waysides or not,  depends on our track circuits.

[draskouasshat]

There are three PTC systems that use GPS as a base.   

Jason

Ill put it to you this way, ptc is supposed to stop a train short of things like other trains,  switches lined improperly, possible broken rails, slide fence locations that have sensed fallen debris etc. These are all things our signal systems warns for. Ptc is added on top of this to essentially take human error out of the equation.
Yes, gps uses the trains location to determine when it needs it apply an emergency brake application if the engineer does not comply with the warnings.
Even if ptc is applied in an area without wayside signals, we still have signal equipment that monitors point position, track circuits etc.

Also, there is only one ptc and everyone uses it. It all works the same and everyone's locomotives works on everyone else's railroad with ptc in use.

Drasko

[draskouasshat]

Also all of the class I railroads own a share of the ONE  company that produces the wayside and locomotive radios. I won't go into detail as to WHY they all do but I'll just say a monopoly had something to do with it.

[learmoia]

And now I see an Acela Train broke a drawbar (supposidly at 125mph).. No derailment.. (I don't think).. based on the photo, looks like a drawbar pin broke.  I'm guessing a backup connection held the train together..

~Ian

[dougnelson]

Looks like it was the cafe car that folded in half.  That should not happen with the US railcar crash standards, but you can imagine what the impact was like.  Passengers must have gone flying into the seats in front of them.

[John]

Just a comment -- and no one should take this personally - We have several professional railroaders here who are giving us good information, including expert first hand knowledge of how these systems (signals, PTC, etc) are supposed to work.  Those systems are completely different of how we operate the model railroads we use, or what we thing we may know. My knowledge of signal systems is based on this book :

https://www.amazon.com/Railroad-Signaling-Brian-Solomon-ebook/dp/B004Q3QKV0/ref=sr_1_2?ie=UTF8&qid=1517950893&sr=8-2&keywords=railroad+signals

and a lot of reading of railroad technical manuals, employee time tables, and discussions with these professionals.

I recommend we learn from what information they have to offer to allow us to become better informed modelers and railfans.   

/>

/>