CSX vs. Amtrak 91 at Cayce, SC

[samfp1943]

Shadow the Cats owner

"...rom what is being said if it was a switch misaligned CSX better reach deep into their pockets..".  They also better bend WAY and I mean WAY over as the FRA STB hell every alphabet agency in the US Government is going to be looking long and hard at them for a very long time on this one.  The crew of that CSX train better be ready also.

 

       This is NOT the first time one of these types of incidents has happened in South Carolina...

  Recall, in Feb 4th, 2005, also in Lexington county, at Granitville,SC.  That time it was  NSRR's trains involved in a similar-type incident. 

 Then in January, 2015, in Allandale,SC, a CSX train entered a plant siding and crashed into standing tank cars...spilling hydrochloric acid and diesel fuel. 

Both incidents, with photos, described in this linked article @

http://www.thestate.com/news/local/article198360629.html

[ccltrains]

From what I am reading on this thread there are two sources of infiormation.  The first is the media which is written by people totally unfamiliar with rail operations.  I tend to disregard 90% of what they say.  The second data source is from railroaders.  They have great knowledge how a railway should be run but have not been on scene to observe in person.  I suggest we wait until we see some "offical" reports from the appropriate investigators.

[BaltACD]

Switches in signalled territory where trains (rule book definition) may CLEAR the Main track and the Main Track speed is designated to be above 20 MPH must be Electric Lock Switches per FRA rules.  If a train is placed in a track without a Electric Lock switch, then the Dispatcher must issue a Speed Restriction of 20 MPH at that switch.

NOW - if that is in fact a normal manual switch, was the Dispatcher informed that locomotives were a part of what was shoved off into the 'siding' thus creating a train.

The audio tapes will tell the tale.

[Overmod]

BaltACD

Switches in signalled territory where trains (rule book definition) may CLEAR the Main track and the Main Track speed is designated to be above 20 MPH must be Electric Lock Switches per FRA rules.

An associated question:  Do 'electric lock' switches still work during a signal suspension?  And if not, does the speed reduction automatically apply, as I would think it surely would, to traffic on the main or otherwise coming up to that switch?

I ask this in part because this accident was almost surely no "20mph" collision.

[tree68]

ccltrains

I suggest we wait until we see some "offical" reports from the appropriate investigators.

Alas, short of updates (which will be less and less frequent), such official reports may be weeks or months in the offing.

I believe most readers here understand that certain factors are very definitely speculation, but that speculation is based on observation and information that has been released.  In addition, most folks qualify their posts.

And, people ask questions about what is "normal."  Those with the knowledge and experience can certain answer those questions.

I'd rather see that than wild speculation based on incorrect theories.

News reports can have a certain amount of useful information.  I like to take a look at the satellite view of incident locations (I do this for fires as well) and a search for local news coverage (vs national outlets) often provides what I need to know to find them.  That was the case here.  

Sometimes what you see in the press images doesn't paint the complete picture.  F'rinstance, there are three tracks visible in many of the press images, which may lead someone to believe that it was at least two-track main - it isn't.

 

[Overmod]

ccltrains

From what I am reading on this thread there are two sources of information.

What you forget is that there is an extensive, and rapidly-growing, third source of information: the various firsthand video documentations showing the facts of the accident.  This is independent of any assumptions, mistakes, or other stupidity of newsworkers, and it provides railroaders with most, if not all, the information they'd need to obtain in firsthand examination of the wreck aftermath.

So in a real sense we've all 'been on scene to observe in person' albeit with some ultimately trifling restrictions on POV.  And it is largely based on those factual observations that most of the discussion so far has proceeded up to now.

I do agree, wholeheartedly, that we should wait for some of the official reports and discussion to be provided before drawing any conclusions about what caused this accident.  But that has nothing to do with the logical process both of elimination and confirmation in assessing the damage and finding likely causes 'in the meantime'.  While some posters here don't always seem to 'get' that difference, I think it is very real and relatively significant.

[Euclid]

What is the distance from the switch to the point of impact?

[Overmod]

Close to, probably a little under, the length of 91's original consist.  You can see the switch location clearly in the views of the last several standing cars, and extrapolate the distance from them to the switch net of any telephoto distortion.  Compare also the relevant views in the drone video.

[F.S. Adams]

Does anyone know if this is Yard Limit territory? I understand it's at the north end of the Cayce yard. If it is yard limit territory and in the absence of signals ATK 91 should have been running at Restricted Speed - prepared to stop short of train, engine, obstrution, switch improperly lined, etc. If it is, still no excuse for leaving the switch open.

[Overmod]

F.S. Adams

Does anyone know if this is Yard Limit territory?

I believe someone knowledgeable has already commented about this earlier in the thread.

[Lithonia Operator]

https://www.cnn.com/2018/02/05/us/positive-train-control-explainer/index.html

Is it accurate that, according to an NTSB official, PTC would have prevented this accident? Does PTC detect switches aligned the wrong way? (Particularly during a signal supension.) Say the siding was empty: would PTC slow a train knowing that no one would ever intend to enter a siding that fast? Would any switch aligned from a main to a siding always trigger PTC actions? Does PTC "know" that (in this case) a train is in that siding? Etc. Etc.

[Murphy Siding]

    

      Am I following this right? The CSX train was parked on the main, through track. The switch was set to send the Amtrak train through the siding to the left of the CSX train? The Amtrak train hit the switch and derailed? As it hit the ground, the locomotive veered to the left so that the right side of the Amtrak locomotive hit the right side of the CSX locomotive?

 

[Lithonia Operator]

Does anyone know how long the CSX crew was off-duty before the crash?

I am assuming toxicology test have now been done on those crew; but was it done soon enough to be accurate enough?

Does anyone know if Amtrak or CSX trains have passed through that location yet since the accident? If not, how were they routed?

[Lithonia Operator]

"Am I following this right? The CSX train was parked on the main, through track. The switch was set to send the Amtrak train through the siding to the left of the CSX train? The Amtrak train hit the switch and derailed? As it hit the ground, the locomotive veered to the left so that the right side of the Amtrak locomotive hit the right side of the CSX locomotive?"

No. The CSX train was in the siding. Amtrak was supposed to hold the main. But the switch was accidentally left lined into the siding, assumedly by the CSX crew.

[zugmann]

BaltACD

NOW - if that is in fact a normal manual switch, was the Dispatcher informed that locomotives were a part of what was shoved off into the 'siding' thus creating a train.

Looks likke an electric lock from the photos.  No target, extra boxes at the lever, and even a relay box next to the ROW.

[Overmod]

Murphy Siding

Am I following this right? The CSX train was parked on the main, through track. The switch was set to send the Amtrak train through the siding to the left of the CSX train? The Amtrak train hit the switch and derailed? As it hit the ground, the locomotive veered to the left so that the right side of the Amtrak locomotive hit the right side of the CSX locomotive?

Since I was part of screwing you up, here is the way I currently see this.

There is one main, the 'middle' track.  CSX shoved back, off the main, into the siding where the collision occurred, with someone 'on the point of the shove' locking the switch so it wouldn't move under the train, then riding the consist back until it was a considerable distance 'clear' of the switch ... the head end and 'point of impact' then being about the length of 91's consist, roughly, distant from the fouling point of the switch.  The CSX train then stopped, and apparently (at this point) no one on the head end went up, unlocked the siding switch, relined it for the main, and locked it again.  Sometime before 91 arrived at this site, the folks on the head end left the area.

91 arrives at the switch around what, to the engineer, was a right-hand curve with view across the 'chord' blocked by trees.  The switch remained set to direct any traffic into the siding, which I now believe the engine and following cars successfully (and somewhat surprisingly) did.  The train then ran down the relatively short length of siding track to a more-or-less bluff collision with the standing train.  Some result of crumple zones and kinematics appears to have folded the nose down in a way that directed the locomotive up and then over to the left, causing substantial damage to the structure of the lead CSX motor, coming to rest where you see it; the first following car suffering a clear compression buckle in its structure, and I suspect the following "V" may be the result of a similar buckle, as there is little else that would produce that effect in an otherwise standing train.  There was likely no 'derailment' up to the point that the Amtrak motor started lifting as a resultant of impact forces and deformation.

I find it highly likely that the train was doing 59mph right up to the point the engineer recognized the open switch, which very well might have been when the engine veered dramatically to the right.  That motion by itself might have delayed application of the brake -- we will have to wait for the release of data from the locomotive EDR, which might not have survived this degree of impact damage, and kinematic reconstructions to see if that might be so.  There was certainly considerable shock reflected back through the train, net of all 'crumple zone' protection in the locomotive nose, which argues to me that the train may still have had considerable way on it at the point of impact.

[Saturnalia]

Euclid

What is the distance from the switch to the point of impact?

 

Point-of-switch to impact was just under 700' as found by the measurement tool on Google Maps, which is reasonably accurate for this purpose. 

As I stated earlier, 60mph is 88 feet per second...simple math tells you there wasn't much time to dump the train, especially if they didn't get to before lurching into the siding, probably knocking the crew off balance, at the least. 

[F.S. Adams]

You didn't answer my question. Is it yard limit territory?

[Saturnalia]

Overmod

 

 

F.S. Adams

Does anyone know if this is Yard Limit territory?

 

 

I believe someone knowledgeable has already commented about this earlier in the thread.

 

This section of track is a single-main, CTC-controlled area. It is not a Controlled Point, so it is in the ABS or "block" signal area. The turnout was equipped with an electric lock. 

What is currently unknown is to what effect the electric lock was being effective with the signal suspension supposedly in progress. Certainly the train crew was rightfully ignoring lineside indications as they operated under EC1 track authority. 

 

I'm going to dump what CSX Rule 266 read as of the 2010 rulebook - no longer in force but gives you an idea of what to expect in a signal suspension, so long as they haven't changed up this rule much. I'd greatly appreciate it if anybody of knowledge of the newer (2015?) book can confirm that this has changed - or not. 

 

Rule 266. Suspension of Signal System

Block Signals and signal rules may be temporarily remove from service by special instructions, dispatcher message or form EC-1

An alternate method of operation must be established and all trains affected notified.

Signals may only be removed from service when authorized by the proper authority, under the following conditions:

1. When a signal system is rednered inoperative by storm or flood,

2. When the signal system is disrupted for other cause and prompt restoration cannot be effected, or

3. When construction work necessitate the signals' temporary removal from service.

Unless otherwise specified, trains must:

a. Approach all absolute signals prepared to stop and not pass these signals without permission of the train dispatcher

b. Approach all public grade crossings that are equipped with automatic grade-crossing warning devices prepared to stop. This must be done unless the warning devices are operating. Trains will procees over such crossings only when it is safe to do so

c. Examine switch points to ensure they are lined and secured before making facing point movement over spring switches

d. Operate switches and derails as prescribed by Rule 104-I, and

e. Stop at drawbridges and railroad crossings at grade and will be governed by rules of special instructions in effect for that particular location.

Rule 104-I has some specifics about Power Switches, Derails and Spring Switches. The switch and derail in this case were not powered, and it was also not a spring switch. 

[oltmannd]

jeffhergert

 

 

oltmannd

Shades of NS Graniteville, SC.   Ugh... All the new rules in place about recording and verifying switch position in dark territory...

So, if signal suspension to install PTC... this could be a PTC caused collision?  

In ABS territory, approaching facing point switch (without circuit controller) while running against the current  meant resticting speed, no?  Wonder why similar not in place when running during signal suspension.

 

 

 

 

No, except in yard limits.  If authorized to run against the current of traffic where not signalled for both directions maximum speeds can be psgr 59mph, frt 49mph. Just like dark territory.  Same with a signal suspension.  Unless otherwise advised you would expect all main track switches to be lined for the main track.

All hand throw switches (and maybe hand operated derails) in signalled territory will have circuit controllers.  (It's the small box you usually see between the rail and the switchstand.  It makes sure the track circuit is shunted when the switch is opened.)  Of course running against the current where not signalled for both directions or signals suspended means there is no signal protection.

 

 

Thanks.  Great explanation.  

[Deggesty]

BigJim

 

 

Deggesty

Larry, I have a question: how long has it been necessary to lock the switch before you can take your key out of the lock?

 

 

Johnny,
Well over tewentysome years, maybe even thirty, at least on the NS. I have no idea how long ago on the CSX.

 

However, it makes absolutely no difference if the switch was locked or not. Sadly, it was lined for the siding!

 

 Thanks, BigJim. So, definitely in 1965 it had not yet been thought a good idea to hold the key in the lock until the lock was closed. Had that been so back then, I could not have helped the crew on their homeward way (about 35 miles yet to go on their 150 mile run). As a non-employee, I would not have had a switch key, even though I knew four station operators and several men in passenger train service--and was recognized by many more men in road service, both freight and passenger.

 

[cefinkjr]

Overmod

 

Firelock76

When I say "...wait for the investigation..." that's strictly in the interest of fairness. ... If you say it's a little ridiculous for an accident report to take two years to finalize I've got no argument with you there.

 

 

You mistake the gist of my comment, alas!  I was merely pointing out that at least one poster in this thread will surely have the thing solved, right down to the engineers' thoughts, motivations, and perhaps alien interference, perhaps in mere days.  With or without any we-don't-need-no-steenkin'-NTSB input.

IOW the comment was not directed at Wayne in any raptor sense whatsoever (nor would any comment of mine likely ever be).  Perhaps I should be more liberal with the emoticons in future.

Yet another example of the need for a font or style expressing sarcasm!  Many people (at least 2) have agreed with me that a sarcasm font or style would be nice to have and that a backward slanting italic might do the trick.  None, however, apparently have the time, skill, talent, or influence to make it happen. 

[oltmannd]

So, what we have is a hand throw switch that is wired into the signal system.  That is, it has a circuit controller so is shunts the track circuit (or equivalent) when reversed. 

However, the signal system is temporarily out of service, so the hand throw switch is then covered by the "belt and suspenders" rules governing hand - throws in dark territory.  The person doing it has to be double checked and forms have to be filled out, including at the dispatching office.

But, since this isn't standard procedure here, it wouldn't be habit, so maybe a mental slip by those putting the train away and the dispatcher?

Or, perhaps that train was put there before the signal suspension, so no record of switch thrown and locked was needed?

[Deggesty]

LithoniaOperator

https://www.cnn.com/2018/02/05/us/positive-train-control-explainer/index.html

Is it accurate that, according to an NTSB official, PTC would have prevented this accident? Does PTC detect switches aligned the wrong way? (Particularly during a signal supension.) Say the siding was empty: would PTC slow a train knowing that no one would ever intend to enter a siding that fast? Would any switch aligned from a main to a siding always trigger PTC actions? Does PTC "know" that (in this case) a train is in that siding? Etc. Etc.

 

Yes, had the PTC installation been complete and in service, there would have been a signal showing STOP. However, two things combined to make the collision possible: the installation was in process, necessitating the temporary inoperation of the signal system, and (far worse) the failure of an employee to realign the switch--MAN FAILURE.

[Euclid]

In a very general sense, it seems to me that there must be around 1,000 feet between the switch and the point of impact.  I assume that it would be normal to make an emergency application of brakes no later than at the switch, or just before it.  

[Deggesty]

Euclid

In a very general sense, it seems to me that there must be around 1,000 feet between the switch and the point of impact.  I assume that it would be normal to make an emergency application of brakes no later than at the switch, or just before it.  

 

Saturnalia posted at 8:47 MDT this morning, "Point-of-switch to impact was just under 700' as found by the measurement tool on Google Maps, which is reasonably accurate for this purpose. 

As I stated earlier, 60mph is 88 feet per second...simple math tells you there wasn't much time to dump the train, especially if they didn't get to before lurching into the siding, probably knocking the crew off balance, at the least. "

Therefore, there was NOT time to take any action, much less effective action.

[Murphy Siding]

Overmod

 

Murphy Siding

Am I following this right? The CSX train was parked on the main, through track. The switch was set to send the Amtrak train through the siding to the left of the CSX train? The Amtrak train hit the switch and derailed? As it hit the ground, the locomotive veered to the left so that the right side of the Amtrak locomotive hit the right side of the CSX locomotive?

 

 

Since I was part of screwing you up, here is the way I currently see this.

 

Thanks for the explanation. I'll take full credit for getting myself screwed up.

[Euclid]

Deggesty

 

Euclid

In a very general sense, it seems to me that there must be around 1,000 feet between the switch and the point of impact.  I assume that it would be normal to make an emergency application of brakes no later than at the switch, or just before it.  

 

 

Saturnalia posted at 8:47 MDT this morning, "Point-of-switch to impact was just under 700' as found by the measurement tool on Google Maps, which is reasonably accurate for this purpose. 

 

As I stated earlier, 60mph is 88 feet per second...simple math tells you there wasn't much time to dump the train, especially if they didn't get to before lurching into the siding, probably knocking the crew off balance, at the least. "

Therefore, there was NOT time to take any action, much less effective action.

 

Thanks for that information.  It would not have given much time for braking. 

[Shadow the Cats owner]

Euclid this would be about what that crew in the cab had time to do in their last 10 seconds of life.  What the F happened and then Holy S when they saw that GE wide nose in front of them.  They had ZERO TIME to even think about grabbing the airbrake lever to dump the brakes.  Why the injuries where so bad and I have a brother in law in that area that knows multiple first responders was the passengers were thrown all over the freaking place in the cars.  There were several compound fractures of legs and arms in those injuries.  The cafe lounge car is scrap metal on site it broke in half like a paper clip.  

[tree68]

cefinkjr

Yet another example of the need for a font or style expressing sarcasm!

Some folks have been known to use the sarcasm "tag" - [sarcasm] at the beginning and/or [/sarcasm] at the end...  

It's not a real tag in HTML or the forum language, so it doesn't screw anything up in that regard.