COLUMBIA, S.C. — The head of the National Transportation Safety Board says a misaligned switch is the immediate reason Amtrak’s southbound Silver Star collided with a parked CSX Transportation freight train early Sunday, killing two people and injuring dozens. Investigators say they’ll now try and find out why the switch was misaligned.
NTSB Chariman Robert L. Sumwalt III told media in a Sunday afternoon news conference that a manual switch was aligned and locked in position for a siding off the CSX Transportation main line near Cayce, S.C. The alignment led the New York City to Miami Silver Star into a CSX autorack train that was parked on a siding.
One locomotive and seven passenger cars collided with the freight train of 34 empty autoracks and two locomotives. The allowable maximum speed for the main line in the area is 59 mph. Eight Amtrak employees were on the train at the time, and 116 of 136 passengers were treated for injuries. There is no information on whether a CSX crew was present.
Officials identified the victims as engineer Michael Kempf, 54, of Savannah, Ga., and conductor Michael Cella, 36, of Jacksonville, Fla.
A CSX representative, in a statement to the media, said that the freight railroad will cooperate with investigators.
“Our sincere condolences go out to the families of the two individuals who passed away following the tragic events that took place in Cayce, S.C., early this morning,” the CSX representative said. “We remain focused on providing assistance and support to those impacted by today’s incident.”
According to Sumwalt and other sources, the CSX train had unloaded vehicles at the Total Distribution Services Inc. transload facility in Dixiana, S.C., on the east side of the main line. TDSI is a CSX subsidiary. The train then moved north on the main, and backed into a siding on the west side of the main, where it tied up.
“Key to this investigation is learning why that was switch was lined that way,” Sumwalt said. “Because the expectation was that the [Silver Star] would be cleared [on the main line].”
Photos of the crash scene show the lead of two CSX locomotives virtually obliterated. The Amtrak locomotive is on its side. An Amfleet coach directly behind it is off the rails, and the fourth and fifth Amfleet coaches are jacknifed off the track. Sumwalt called the damage “catastrophic.”
Sumwalt said NTSB investigators, assisted by Federal Railroad Administration, and an FBI forensics team will be on the scene for about a week.
“Our goal is to not only find out what happened but why it happened, so we can prevent it from happening again,” Sumwalt said.
The collision at Cayce is southwest of Columbia, the South Carolina state capital. On this stretch of track, the Silver Star rides on CSX’s Columbia Subdivision, formerly the Seaboard Air Line main line.
In a statement on the Amtrak website, passenger railroad officials said that the Silver Star trains 91 and 92 will detour between Hamlet, N.C., and Savannah, Ga., until further notice. Officials say the Auto Train, Palmetto, and Silver Meteors are operating normally.
— UPDATED: New lead sentence; detail of the accidents; and comments from CSX Transportation, Amtrak, and the NTSB. Feb. 4, 2018, 6:46 p.m. Central time.
The NTSB has issued an Urgent Safety Recommendation:
“To the Federal Railroad Administration:
Issue an Emergency Order directing railroads to require that when signal suspensions are in effect and a switch has been reported relined for a main track, the next train or locomotive to pass the location must approach the switch location at restricted speed. After the switch position is verified, the train crew must report to the dispatcher that the switch is correctly lined for the main track before trains are permitted to operate at maximum-authorized speed. (R-18-005) (Urgent)”
https://www.ntsb.gov/investigations/AccidentReports/Reports/RSR1801.pdf?utm_source=news&utm_medium=email&utm_content=railroad
Hard to believe this was not SOP.
/Mr Lynn
Yes, “try to …” is the correct form, whereas “try and …” is a careless misuse and, unfortunately, a common one.
Was this a track switch thrown by the crew of the Autorack train and forgotten? Was this a two man crew and they didn’t check each other? Was this remote switch and lock? Where there any kid of block signals that would have slowed or stopped the train interlocked with the switch? Looking at pictures you could pretty well guess what happened, either misaligned or faulty switch.
“Investigators say they’ll now try and find out why the switch was misaligned.”
Investigators will try, and also they will find out why the switch was misaligned? Are trying and finding out two separate activities, or should that read “try TO find out”?
Oh well, since this is the USA, where we don’t use the Queen’s English, “it don’t make no never mind.”
“Investigators say they’ll now try and find out why the switch was misaligned”
Or
“Investigators say they’ll now try to find out why the switch was misaligned”?
Hard facts and polite, reasonable, and respectful discussion, such as the comment by Michael Walker, are much more desirable and appropriate than the snarky, insulting, and unfounded remarks by several other contributors, some of whom appear to be ill-informed about the railway industry and operations, who often do not bother to get themselves straight on the applicable facts before spouting off with their illogical or irrelevant personal opinions or speculations.
Having re-read my post, I should clarify that the head-on collision I refer to I omitted to add “on a single line”. Such are our safety systems on such lines (widely laughed at by US observers) that the one previous to that was in 1921! The last head-on collision on a multiple-track line in the UK was at Ladbroke Grove in 1998 which was the result of a SPAD due to inadequate training as a result of which the entire training regime in the UK was overhauled.
Re- the post from James Hall on Feb 5 regarding my earlier comments.
The “accidents” referred to in the UK are of a totally different magnitude. These are mainly minor derailments in yards or fatalities caused by members of the public choosing to use the railways as a means of taking their own lives which is an all to common occurrence despite having probably the best fenced railways in the world. Under the UK system, all incidents, however “minor”, are required to be reported to the Rail Accident Investigation Branch (RAIB) and it is they who decide whether it justifies an investigation or not.
The last serious passenger train derailment in the UK resulting in a loss of life was that of a Virgin Pendolino at Grayrigg in February 2007. The last incident of a head-on collision on a UK rail line was in 1994. True, there have been minor incidents such as a train derailing leaving a siding outside Paddington due to a SPAD a couple of years ago or the derailment of a Yeoman stone train leaving the quarry last year but these pail into insignificance compared to the accident record not only of Amtrak but of the US industry in general – it seems barely a week goes by on this website without a report of a serious collision somewhere.
When you consider that we in the UK (and likewise in most European countries) run more trains in a day on a 10,000 mile system than the entire US network does in a week then I don’t think our record can be held to ridicule.
As someone actively involved in ensuring operations are carried out safely over here, I object to our safety record – widely acknowledged as being the best in the world – being compared with yours.
the Union Pacific has hand throw switches from the main track to it’s yard tracks. When the switch is open it activate a signals in that block. The switch has a red disc that turns when switch is thrown, showing that the switch is open (lined to the yard tracks). When the switch is lined back for the main the lever must be pushed down all the way for the lock to be put on the lever. The red disc can be seen from a distance an added safety for when the signal system is not working.
One last time (for me anyway) I am still shaking my head at some of the comments in here. Regarding the electric lock switch, yes Mr. Farkas it is a hand throw switch as I stated way back. In regards to having a box next to it that requires throwing a lever to unlock the switch, that is more of a model variation than a specific requirement. Yes, all electric lock switches have to have mechanisms in them to prevent them from being opened when they shouldn’t be, but there are models out there that have the box, and some do not. When I started in 94, the electric locks we used had a foot pedal on them. You pulled the switch lock out of the hasp, and if a light on top of the switch hardware lit up the switch was unlocked. If it did not light up, the switch was not unlocked. In addition, as I said before, the switch wont even unlock (in your case, the lever in the box wont throw) unless certain parameters have been met.
Regarding all of this restricted speed talk because the signal system was suspended and the electric lock was not “protected” (someone else’s words, not mine), would someone please quote me the rule number because I sure as heck can’t find it in my GCOR. Under normal circumstances, if a train has to be talked by a red signal, yes restricted speed applies. But the signal system at the time was suspended, and that is NOT a normal circumstance since basically, with a couple of exceptions, the signals do not “exist” (for lack of a better description). Under a signal suspension the trains operate under a different set of parameters. When operating under a signal suspension, trains follow the rules that apply to operating in non signaled (dark) territory, hence the 59/49mph maximums, provided of course the track is good for those speeds. Do trains running in dark territory run prepared to stop short of every switch? That being said, there are certain situations where a train operating under a signal suspension may be required to stop short of a signal or a switch, but it has nothing to do with an electric lock switch.
I have been through my share of these over the years and will say there is a lot going on when proceeding through one of these. It’s also pretty intimidating to come up to a red absolute and pass it at speed, but you treat it like it is not there.
Why was the switch left in the open position? That is for the NTSB to find out. My heart goes out to my two “Railroad Brothers” that lost their lives.
I agree with Mr. Farkas! In a previous post, I questioned the speed situation. Observing the damage, I’m not all that sure that the speed was max authorized 59 given the overall damage. “Generally, as you noted, when the CTC goes down everything red/occupied and in these situations trains generally have to stop, get authority from the dispatcher to pass that red signal and proceed at a reduced speed in the distance necessary to stop the train generally no greater than the engineer’s line of sight looking out for obstructions, etc. That’s more likely than not, not going to exceed 15 mph @ 3AM double oh pitch black dark thirty.
The Amtrak locomotive may have derailed slightly before the impact possibly as it went through the open diverging switch, causing a right front/right front impact or the impact occurring on a sweeping curve may have released some of the force, not to diminish the severe forces seen here, causing the Amtrak locomotive to flip on it’s side and come to rest 2 car lengths from the point of impact. Either it was going slow…er or the impact nearly stopped the forward motion.
The damage here reminds me of 2 trains colliding because one was past the derail or not all the way into or just coming out of a siding. In addition, we’ve all seen those “cornfield meets” where 2 trains meet at speed coupler to coupler on a straight track. UGH! The Amtrak train may have been looking like an accordion as opposed to the I believe 1 accordion and a couple of mid car buckles like the damage on the car just to the rear of the power which may reinforce the head-on scenario. Maybe the “tight-lock” couplers did their job.
All of this is pure conjecture accompanied with the appropriate, “More likely than not’s”, “possibles” and could’ve, should’ve, and would’ve, etc., much like the final accident report in a couple of years. Time will tell. In the meantime it gives us plenty of “fodder” for discussion.
Steven: In your expert judgement and experience, about how fast do you think the train was going upon impact?
Thanks for the clarification, Steven.
Regarding rail passenger safety and efficiency…Amtrak’s days are numbered! As a corporate governance almost totally dependent on Federal Railroad Administration grants to operate, professionalism in transit operations is very bad. Recently; the media both industrial magazines and the public press have disclosed that Amtrak does not keep adequate accounting. What are they spending their tax based dollars on?
Just in the last few months Amtrak has dumped a special congressional train loaded with legislative members. Thankfully only a few were seriously injured, but sympathy to the family of the truck driver killed on the crossing. Media reported that the crossing gates were not functioning!
Incredible crash in Washington state with the “Cascades” would have been prevented with full attention by the engineer. Similar to the Philadelphia crash 2 years ago…for sure! Incompetent operations. Now the latest wreck on CSX. Switch gear not in place, too bad for the passengers. Here’s a clue rail fans…that train was probably in what is called a “dark area”. That is when a section of distance on rail is not covered by signals. That’s why this wreck happened in the first place. Amtrak should not operate in area were there is these “dark zones”! Again; Amtrak’s route planning is suspect.
Amtrak’s days are numbered! Only a matter of time till our business minded President Trump axes this ridiculous operation that has been wasting taxpayer money and killing rail passengers since 1971. Trump understands business he has a MBA from the Wharton School University of Pennsylvania, one of the best business schools! He knows what transportation waste is.
Philip, I did not say that the dispatcher’s board would show as one big occupancy. I said it MAY just show one big occupancy. Without talking to a dispatcher, I don’t know what he sees. My point was that during a suspension he has no control over them. And an electric lock switch IS a hand throw switch. The meaning of “electric lock” has nothing to do with whether or not there is a physical lock on the switch. All mainline switches have (or should have) a lock on them. The “Electric Lock” term means that the switch cannot be opened randomly in signalled territory without certain conditions first being met (ie: the dispatcher does not have the signals lined for a move, no train approaching, etc).
I saw a story that noted (correctly or incorrectly) that “the lock was on the switch (stand)”). One could interupt that to mean the lock was visible as opposed to being an electric lock. That would indicate to an observer that the switch was likely the old fashioned kind that somebody has to get out, unlock the stand, twist the switch stand and lock it back up.
Mr. Bauer noted that there are protocols in place when the CTC system is down and the dispatcher’s board shows 1 big occupancy. That means all signals are red or dark. Plus, the dispatcher has no control over any automatic switches, which means the train needing to move a switch would have to stop, get permission to place the switch in “hand”, moved, and reset to auto operation. Or a “switch tender” or manager would have to be at every switch. With all of the austerity programs in place I doubt if the CSX would spring for all those people. Of course, managers are salaried, they can work anytime or for as long as they’re needed. LOL Wasn’t there a big war about working for free in this country?
Respectfully, I wonder if the rules are universal in North America or depend on the railroad as to how the train would proceed, likely after stopping? Would the train proceed at a reduced speed or max authorized speed or some other speed in between dark or red signals with authority from the dispatcher or corridor manager? Can or would that authority be a blanket authority given between mile markers or dark or red signals? Say, milepost 36 and 50 which would be control points, then have to stop, or just continue to get authority on the move like it was a train order territory? An interesting situation.
Re this, from a James Hall, yesterday 5th. Feb:
The record in the United Kingdom is no better — nor, honestly, much worse — that that in North America (your comment on accidents per decade is just wrong — there have been at least four, at least one fatal, in the past month alone).
Could you please let us know where they were? Many thanks.
John Corkill – old 578 operater at Ohio Railway Museum
I’m not sure what happened to the Amfleet car that folded in the middle.
Would that be a failure of the American rule that a passenger car must
withstand 800,000 pounds of force, end to end, without telescoping or
collapsing ? If the colliding speed were 50 MPH, one assumes the NTSB
could calculate the stress on that car from the sleepers and baggage
car behind it. As the man said, “good thing there were no midnight snackers
at the bar.”
It’s amazing to me, in all of the comments on here, how many people are still referring to this as a non signalled railroad, how the dispatcher should have known the switch was open, etc. And there also seems to be some confusion over the type of switch it was. My understanding is it was an electric lock switch, which is a totally different animal from a power switch. The electric lock can only be opened when certain parameters have been met but the dispatcher does not throw it.
It is a signalled territory, but the signal system was under suspension at the time! It was not an insignalled line. The few posters in here who are, or were, actual RAILROADERS versus just knowing a lot about trains (there is a difference), know that you just dont suspend a signal system without protocols in place to move trains during the suspension. I went through three or four of.them back about 10 yrs back when the outfit I work for was installing new signals on a line that was previously just governed by cab signals, in addition to when new control points were put in on a section of the track I run on just a few years ago.
When the system is down, the disp has no control over ANYTHING. In fact, his whole board in the out of service segment may just show as one big occupancy. I would have to talk to one of the guys and ask what he sees on his board. That being said, even though he does not have control over the switches, that does not mean the trains just go through there willy nilly and take their chances. It would seem to me in this case that for whatever reason the switch was not lined back before the next train was cleared through. As I have seen in my career, sometimes things aren’t always what they seem. But that is for the investigators to figure out.
One last thought: if the PTC had been installed on this line (or any other for that matter) and the signal system would have been suspended for another reason, would the PTC have been useless? Since it reads off of the signals and switches, I wonder if the whole PTC screen would show a red area through the whole suspension and not let the train move. We don’t have it yet where I run but it’s coming, so.I don’t have an answer to that.
I have been at this 24yrs, and hopefully have only 6yrs and 8 months of this left, but who’s counting.
Great Britain/Scotland: Quintinshill/Gretna Green, May 22, 1915 on the Caledonian Railway. 227 dead, including 214 men of the Royal Scots. “I forgot.” The Brits were furious when it was pointed out that Quintinshill couldn’t have happened on an American railroad–we do our wrecks differently.
It appears that a simple signal control circuit on the siding switch would have prevented this accident. How many main line switches in the US have this exposure? An accident like this can have many causes ranging from fatigued or impaired crew members to outright vandalism or terrorism. Any nonsignaled railroad that permits 59 mph speeds over possible open switches needs a safety review.
PTC kills? Like the Old Colony RR 1895 collision (23 dead) during the weekend they were running unsignalled during the changeover from left-hand to right-hand running.
Electric Lock Switch! This is a hand throw switch with a railroad lock on it, there is also a Black or Silver box close to the hand throw switch, you have to unlock this box and throw the lever in the box to UNLOCK before you can throw the switch, you can’t throw the track switch unless this lever in the box is UNLOCKED. You can’t lock this box in a open or deverge position, you can lock the switch open but the box will not lock until lever is lined in the box , in fact you can’t even close the box door unless the lever is thrown So this area is occupied by the CSX train. Why is AMTRAK coming down the line at 59 MPH? Dispatcher Radio contact? Everything should be RED, OCCUPIED. I’m not blaming the AMTRAK Crew, I think this is the CSX Dispatchers mess.
Although Freddie Frailey has gotten everyone all riled up recently, anyone who thinks Ron Batory will be some kind of FRA white knight to solve all our troubles really doesn’t understand the concept of “regulatory capture” very well.
In response to Mr. Hume’s prior commentary, I disagree with turning this unfortunate event into a political football, but I would like to clarify his misrepresentations on several fronts. The territory in which this accident occurred was indeed signaled territory, with the signal system suspended due to PTC installation by a private corporation, namely CSX. We can debate the safety rules and protocol in place, but that would rest solely under the responsibility of the CSX operating rules. I would like to make clear in this particular incident, the Amtrak engineer and conductor who perished in the accident most likely did absolutely nothing wrong, and were following the operating rules of the railroad they were operating on (CSX)…there is no need to troll fake news assumptions concerning the incident and tarnish their reputations. I would imagine that at the end of the day, the blame will probably fall on the failure of the CSX crew for failing the restore the switch and leaving the location after reporting that they were in the clear, but time will tell when the investigation is complete.
In addition, the grade crossing incident with the GOP contingent on board last week also occurred on another railroad, Buckingham Branch Railroad, and Amtrak was not responsible for maintenance of the grade crossing protection, as they neither own nor maintain the tracks or right of way.
My commentary comes from the point of view of an actual railroader, and will gladly clarify any questions concerning signal systems and dark territory, Mr. Hume, if you’re interested in how things really work. Responsibility is spread out over a great many people, but it was most likely not the deceased Amtrak operating crew in this particular case.