Amtrak Wreck in Philadelphia

[CMStPnP]

MrLynn

I wonder if it could be relevant that to accelerate the ACS-64, you push the combined throttle/dynamic brake lever forward, but the Acela has a throttle you pull back to accelerate:

The train on engineer Bostian's previous run was an Acela.

/Mr Lynn

 

That just might have something to do with it.    Also raises the question why Amtrak does not have a standard or standards concerning engine dashboard controls between trainsets?

[Euclid]

schlimm

 

Euclid

To me, Sumwalt’s objecting reaction to adding the word “likely” suggests that this conclusion about the radio causing distraction is an agenda conclusion on the part of the NTSB.  They appear to be working too hard to blame it on distraction caused by a radio transmission. I think it is far more likely that the engineer was distracted by a rock hitting his windshield.  How do we know that didn’t happen?

 

 

You are inferring far too much, perhaps because you do not understand deduction from various pieces of empirical evidence vs reporting ALL the evidence gathered (likely in some appendices).

 

Schlimm,

I do understand deduction from various pieces of empirical evidence.  I assume that is what the NTSB is doing, however, I must question their deductions.  They have now changed one of their deductions that I disagreed with, and so now I no longer disagree with it.

But I am still fascinated by their deduction that the engineer lost situational awareness due to distraction rather than incapacitation.  They say the engineer began accelerating using the procedure that he normally used.  They know that because the engineer told them that he remembered doing that.  The engineer then told the NTSB that he has no memory from that point where he executed his procedure to accelerate.  In other words, the engineer told them that was the point where his memory loss began. 

From there, the NTSB concludes that the engineer’s failure to slow for the curve was due to him thinking he was past the curve.  Maybe that is a reasonable deduction.  Earlier here, I asked why the engineer would have executed his acceleration routine if he knew he would have to slow for the curve almost immediately after accelerating.  I don’t know the exact distances involved, but maybe the speed restriction for the curve was so close that there was not room for the engineer to reach his intended speed.  Or maybe his intended speed was too high for the territory he was running in, so those factors, if true, would indeed indicate that he was lost.  The answer to those points would resolve whether the engineer was lost or incapacitated, and maybe the NTSB has addressed those points and found that they support their conclusion that the engineer was lost.    

Also, no matter where the engineer said he lost his memory, that is claim is unreliable.  He may have lost his memory from the trauma of the derailment, and lost it from that point, extending to some point earlier; but while operating the train within that span prior to the derailment, his memory might have been functioning normally.  Or he may have lost his memory due to some unexplained cause, and lost it exactly at the threshold where he now says he cannot remember anything after.  He also may have been able to operate the train normally during a phase where he was recording no memory. 

It is also possible that the engineer has not lost any memory as he claims.  If he has lost memory from the beginning of acceleration and if the NTSB is correct about the distraction causing the loss of situational awareness; then the two different and unrelated factors coincidentally apply to the same period of the engineer’s experience in approaching the curve.      

 

[wanswheel]

It's probable I guessed the SEPTA train was likely on his mind.  

http://cs.trains.com/trn/f/743/t/254126.aspx?page=2

 

[Euclid]

wanswheel

 

Euclid

I am not aware of the use of the word, “probable,” but perhaps that is used in the broadcast and not in the text, as is the case with the word “likely.” 

 

 

 

Please click the link to the abstract again. Today it says,

“The National Transportation Safety Board determines that the probable cause of the accident was the engineer’s acceleration to 106 miles per hour as he entered a curve with a 50 mile per hour speed restriction, due to his loss of situational awareness likely because his attention was diverted to an emergency situation with another train. Contributing to the accident was the lack of a positive train control system.”

http://www.ntsb.gov/news/events/Documents/DCA15MR010_Abstract.pdf

 

Wanswheel,

That is interesting.  With their change of wording of the probable cause by that addition of the words “probable” and “likely,” they are no longer stating a conclusion as though it is a fact, so I now see nothing to disagree with.  However, they did not change the wording in the section above that with the heading, Crewmember situational awareness and management of multiple tasks.

There they say:

“He lost his situational awareness because his attention was diverted to an emergency situation with a nearby SEPTA train that had made an emergency stop after being struck by a projectile.”

[schlimm]

Euclid

To me, Sumwalt’s objecting reaction to adding the word “likely” suggests that this conclusion about the radio causing distraction is an agenda conclusion on the part of the NTSB.  They appear to be working too hard to blame it on distraction caused by a radio transmission. I think it is far more likely that the engineer was distracted by a rock hitting his windshield.  How do we know that didn’t happen?

You are inferring far too much, perhaps because you do not understand deduction from various pieces of empirical evidence vs reporting ALL the evidence gathered (likely in some appendices).

[wanswheel]

Euclid

I am not aware of the use of the word, “probable,” but perhaps that is used in the broadcast and not in the text, as is the case with the word “likely.” 

Please click the link to the abstract again. Today it says,

“The National Transportation Safety Board determines that the probable cause of the accident was the engineer’s acceleration to 106 miles per hour as he entered a curve with a 50 mile per hour speed restriction, due to his loss of situational awareness likely because his attention was diverted to an emergency situation with another train. Contributing to the accident was the lack of a positive train control system.”

http://www.ntsb.gov/news/events/Documents/DCA15MR010_Abstract.pdf

[Euclid]

 

Wanswheel,

Okay, I see. That is most interesting about your mention that Robert Sumwalt objected to adding the word “likely’ because the word “probable” is used and it means the same as “likely.”  I am not aware of the use of the word, “probable,” but perhaps that is used in the broadcast and not in the text, as is the case with the word “likely.” 

To me, Sumwalt’s objecting reaction to adding the word “likely” suggests that this conclusion about the radio causing distraction is an agenda conclusion on the part of the NTSB.  They appear to be working too hard to blame it on distraction caused by a radio transmission.

I think it is far more likely that the engineer was distracted by a rock hitting his windshield.  How do we know that didn’t happen?

 

[wanswheel]

Euclid, the abstract published yesterday was written before the board met. Expect that second sentence to have ‘likely’ between ‘awareness’ and ‘because’ in the final report.

The webcast, fast forward to about 2:14.

http://ntsb.capitolconnection.org/051716/ntsb_archive_flv.htm

http://ntsb.capitolconnection.org/#

[Euclid]

wanswheel

This is why Member Weener wanted the word “likely” inserted in the probable cause, to reinforce the fact they didn’t actually know it was the SEPTA distraction that truly caused Bostian to lose his situational.  Member Sumwalt was a little bit miffed, saying, essentially, it’s redundant, “likely” means about the same thing as “probable” anyhow, but then he said it was a “friendly amendment,” so no problem.

Wanswheel,

Apparently, you are referring to the live broadcast.  I have not seen that, so I am only going by the text from the link you provided above, and focusing on two sentences from that text as shown below.

Maybe there is a larger context where the word “probable” applies in blanket fashion to every speculative conclusion by the NTSB, but in considering the two following sentences, there is no use of the word “probable.” 

Sentence #1

"After evaluating the circumstances of the accident, the NTSB found that the most likely reason the engineer failed to slow for the curve was he believed he was beyond the curve where the authorized speed was 110 mph, because of his loss of situational awareness."

Sentence #2

"He lost his situational awareness because his attention was diverted to an emergency situation with a nearby SEPTA train that had made an emergency stop after being struck by a projectile."

 

The word “likely” is used in the first sentence to support ONLY the conclusion that the engineer lost situational awareness.  Indeed, the loss of situational awareness seems obvious unless the engineer was motivated by something extraordinary such as a suicide attempt. So the word “likely” is entirely reasonable for the first sentence.

However, there is nothing about the conclusion of sentence #1 that confirms the truth of sentence #2.  There many different plausible explanations for the loss of situational awareness besides being distracted by the radio transmission about the rock throwing. 

So just because sentence #1 is qualified by the word “likely,” that does not mean that the word “likely” also applies to sentence #2.  Therefore, the NTSB offers no qualifier to sentence #2.  Instead, it is asserted purely as a fact to explain the loss of situational awareness. 

 

 

[Euclid]

dehusman

 

Euclid

Here's why: They say it can’t be incapacitation (other than distraction) because the engineer accelerated his train according to normal procedure. But what if he became incapacitated AFTER he began his normal procedure of acceleration? How do they rule out that possibility?

They are painting with a broad brush and you are splitting hairs. 

Distraction is not the same as incapacitation.  Distraction means you lost focus or changed focus, incapacitation means you were unable, you lacked the capacity.  From the evidence they have, they determined that he had the capacity but not the focus.  To reach a different conclusion requires different or additonal evidence.  The NTSB couldn't find it.  If you have it let them know.  Otherwise they (as well as the rest of us) have to go with the evidence that there is.

I agree with what you say about distraction being different than incapacitation.  I said exactly that in my post that you extracted the quote from.  I also explained the defect in their reasoning that concludes that situational awareness was caused by distraction rather than incapcitation.  The reason for that conclusion does not hold water.  It does not preclude incapacitation, as they conclude it does. 

[dehusman]

Euclid

Here's why: They say it can’t be incapacitation (other than distraction) because the engineer accelerated his train according to normal procedure. But what if he became incapacitated AFTER he began his normal procedure of acceleration? How do they rule out that possibility?

They are painting with a broad brush and you are splitting hairs. 

Distraction is not the same as incapacitation.  Distraction means you lost focus or changed focus, incapacitation means you were unable, you lacked the capacity.  From the evidence they have, they determined that he had the capacity but not the focus.  To reach a different conclusion requires different or additonal evidence.  The NTSB couldn't find it.  If you have it let them know.  Otherwise they (as well as the rest of us) have to go with the evidence that there is.

The only thing I'm suprised about is they didn't come out stronger for inward facing cameras.

[wanswheel]

This is why Member Weener wanted the word “likely” inserted in the probable cause, to reinforce the fact they didn’t actually know it was the SEPTA distraction that truly caused Bostian to lose his situational.  Member Sumwalt was a little bit miffed, saying, essentially, it’s redundant, “likely” means about the same thing as “probable” anyhow, but then he said it was a “friendly amendment,” so no problem.

 

[Euclid]

 

dehusman

 

Euclid

But how can the NTSB assert as a matter of fact that the reason he lost situational awareness was because of the radio transmission regarding the rock throwing incident? What is the evidence for that? Have they answered that question? I have not seen it in the news reports. Did the engineer tell the NTSB that he believed he lost situational awareness due to his reaction to the radio transmission about the rock throwing?

Remember the NTSB has determined "probable cause".  Since "situational awareness" is basically something in somebody's head, unless the person tells the NTSB exactly what they were thinking about at the moment the wrong action was taken, there is no way that the NTSB can determine EXACTLY what the person was thinking at any given time.  Thoughts don't leave a paper trail.  From the engineer's testimony he said he remembers the rock throwing incident.  It was the only unusual thing that has been documented that was happening at the time, it is probable that was the cause of the distraction. 

 

That is my point.  The NTSB does not know what caused the situational awareness.  Yet they assert that as a known fact.  Although, maybe it is not an assertion.  It’s hard to tell because they say one thing and then restate it in a slightly modified way that allows an apparent assertion to be interpreted as simply laying something on the table as a "what if."

Here is another point:  In order for their “distraction” explanation of loss of situational awareness to work, they have to rule out the other possibility of the engineer having been “incapacitated.”  Yet distraction is one form of incapacitation.  So what they are ruling out is incapacitation from a cause other than distraction, such as a micro sleep phase or some other type of loss of consciousness.  However, the method of logic used by the NTSB to replace incapacitation with distraction strikes me as being faulty.

Here's why:  They say it can’t be incapacitation (other than distraction) because the engineer accelerated his train according to normal procedure.  But what if he became incapacitated AFTER he began his normal procedure of acceleration?  How do they rule out that possibility?

That explanation cannot work for the NTSB because it would rule out the timing for the loss of situational awareness to have been caused by the radio conversation, and that it the explanation they want. 

I don’t think you can determine what distracted a person simply by selecting the seemingly most important event encountered by that person in the timeframe where the distraction is thought to have occurred.        

 

[dehusman]

Euclid

But how can the NTSB assert as a matter of fact that the reason he lost situational awareness was because of the radio transmission regarding the rock throwing incident? What is the evidence for that? Have they answered that question? I have not seen it in the news reports. Did the engineer tell the NTSB that he believed he lost situational awareness due to his reaction to the radio transmission about the rock throwing?

Remember the NTSB has determined "probable cause".  Since "situational awareness" is basically something in somebody's head, unless the person tells the NTSB exactly what they were thinking about at the moment the wrong action was taken, there is no way that the NTSB can determine EXACTLY what the person was thinking at any given time.  Thoughts don't leave a paper trail.  From the engineer's testimony he said he remembers the rock throwing incident.  It was the only unusual thing that has been documented that was happening at the time, it is probable that was the cause of the distraction. 

I am not suprised at the finding.  Once they ruled out mechanical and track, it looked like a situational awareness incident.  Somebody doing normal stuff at he wrong place and time.

I just wish the national new media would actually take the time to read the report and stop calling the engineer a "conductor". 

[BigJim]

Wizlish

and YES, it's highly likely that a reflective 'permanent speed restriction' warning sign would have prevented the accident completely.

They don't have curve speed boards? Why not???

[RME]

wanswheel

The NTSB found that if the passenger car windows had remained intact and secured in the cars, some passengers would not have been ejected and would likely have survived the accident. Further, the passengers were not protected from serious injuries resulting from being thrown from their seats when the cars overturned. The NTSB concluded that the current passenger equipment safety standards are not adequate.

I've been staying out of the latest round of speculation, but here is an item that deserves some further discussion.

It occurs to me that the discussion of windows breaking out of passenger cars and allowing ejection or rollover injury to passengers is a very old issue -- but so is the stated desire to have safety windows that allow ease of egress after an accident.  How would the NTSB or other agency reconcile these two considerations.  More to the specific point, how would the NTSB propose to implement workable technical car improvements (that are maintainable and have reasonable customer and 'stakeholder' satisfaction) that would allow the various types of car window to survive overturning accidents?

The 'being thrown from their seats' sounds to me like a code word for mandating seat restraints, probably seat belts.  Now, I've seen it discussed here and elsewhere that belts on a train are not particularly safe or productive, but they sure are relatively expensive to provide, let alone retrofit with full required design integrity, and are sure going to be perceived as irritating by a large cross-section of passengers.  (Compare the arguments for seat belts on school buses, another key 'regulator' item for attention from time to time.)

This looks like another set of T. Bella Dimh-Ouidh priorities to me (I confess to being in an uncharitable mood this morning, still a bit irritated over the idea that PTC can be conflated with automatic train stop and considered the primary proximate cause of the accident.)

Who has ideas about making effective (we can get to cost-effective later in the 'design process') glazing integrity systems for -- well, let's stick with the immediate new-build priority, Viewliners?  Who has thoughts about preventing or minimizing passengers suffering trauma from seat ejection in unanticipated rollovers?  I think there is at least the possibility for some intelligent brainstorming from this community in both respects.

[tree68]

Given that most of us have about the same knowledge of the incident and the information that's been released, how can any but NTSB answer that?  

At this point, it's time to just sit back and wait.  Hysterical speculation will get us nothing.

[Euclid]

wanswheel

Excerpt from NTSB abstract

He lost his situational awareness because his attention was diverted to an emergency situation with a nearby SEPTA train that had made an emergency stop after being struck by a projectile.

 

I think that it's reasonable to assume that the engineer lost situational awareness.  He said he has lost his memory of the events leading to the derailment. 

 

But how can the NTSB assert as a matter of fact that the reason he lost situational awareness was because of the radio transmission regarding the rock throwing incident?  What is the evidence for that?  Have they answered that question?  I have not seen it in the news reports.    

Did the engineer tell the NTSB that he believed he lost situational awareness due to his reaction to the radio transmission about the rock throwing?

 

[wanswheel]

Excerpt from NTSB abstract

http://www.ntsb.gov/news/events/Documents/DCA15MR010_Abstract.pdf

 

Crewmember situational awareness and management of multiple tasks. The NTSB found that the Amtrak engineer accelerated his train to a high rate of speed in a manner consistent with how he habitually manipulated the controls when accelerating to a target speed, suggesting that he was actively operating the train rather than incapacitated moments before the accident. However, he accelerated to 106 mph without slowing the train for the curve at Frankford Junction, where the speed was restricted to 50 mph. After evaluating the circumstances of the accident, the NTSB found that the most likely reason the engineer failed to slow for the curve was he believed he was beyond the curve where the authorized speed was 110 mph, because of his loss of situational awareness. He lost his situational awareness because his attention was diverted to an emergency situation with a nearby SEPTA train that had made an emergency stop after being struck by a projectile. This type of situation could be addressed by better crewmember training that focuses on preventative strategies for situations that could divert crewmember attention.

 

Positive train control. In the accident area, positive train control had not yet been implemented at the time of the accident, but it has since been implemented. The NTSB found that the accident could have been avoided if positive train control or another control system had been in place to enforce the permanent speed restriction of 50 mph at the Franklin Junction curve.

Passenger railcar window systems and occupant protection. The NTSB found that if the passenger car windows had remained intact and secured in the cars, some passengers would not have been ejected and would likely have survived the accident. Further, the passengers were not protected from serious injuries resulting from being thrown from their seats when the cars overturned. The NTSB concluded that the current passenger equipment safety standards are not adequate.

[blue streak 1]

NTSB certainly did not address very well  the lack of ATC limitations on the curve going northbound.  Understand it was in service by time route restored ?  Would PTC requirements have been implemented by just a more robust ATC at a much lower cost ?

 

[BaltACD]

I 'believe' I read somewhere that the throttle on Acela's and the ACS-Sprinters work opposite of each other.  For 'faster' one gets pushed ahead, in the other 'faster' requires the throttle to be pulled back.

Engineer Bastian operated an Acela to DC and was operating an ACS-Sprinter on the 188 trip. 

Situational awareness, stress, fear.

I could be mistaken - if both throttles operate in the same manner - this post is null and void.

[wanswheel]

The webcast of the NTSB meeting was fascinating.  Member Weener insisted the word “likely” be inserted before the word “because,” as to the Septa radio transmissions, which distracted the engineer, which caused him to lose situational awareness, which caused him to accelerate into the curve, which caused the train to derail.

The Vice-Chairlady proposed an amendment that lack of PTC be the probable cause, instead of a contributing cause, which was voted down. She voted against the final probable cause, which passed 3-1.

 

[schlimm]

That is your unsupported conclusion.  They simply did not yet publish the reasons and evidence supporting their findings.  That does not mean they have none or that it is "just speculation."

[Euclid]

The news now is that the engineer was distracted by the radio conversation about rock throwing.  Apparently that is just speculation on the part of the NTSB.  They offer no reason for for the assumption. 

[wanswheel]

NTSB webcast tomorrow morning

http://www.ntsb.gov/news/events/Pages/2016-Amtrak-BMG.aspx

 

[Wizlish]

Euclid

I think it was at least a mistake to overlook the fact that the reasoning they applied to protecting the southbound route also applied to the northbound route; perhaps equal to the mistake the engineer made if he forgot where he was.

I was looking at PDN's comment on one of Scanoes' ten90 blogs this morning which states the situation very well.  The curves before and after the critical Frankford Junction area are said to be similar right-handers of about the same radius, perhaps without visible landmarks at night.  Common practice was to accelerate to high speed after exiting the 'second' one going northbound; I see it as very likely Bostian mistook the first for the second -- none of the other data from the alerter system, etc. backs up any 'accidental' acceleration to the extent reached in those last 78 seconds or so.

YES, an overspeed protection system would have caught this, and YES, it's highly likely that a reflective 'permanent speed restriction' warning sign would have prevented the accident completely.  (Schanoes mentioned in a different context that Metro North instead of putting up a fixed sign at the site of the Spuyten Duyvil catastrophe chose to inject a 30 mph code signal into the rail at that point, which is not at all comparable...)

[Euclid]

Norm48327

 

Euclid

 

blue streak 1

There is something Amtrak needs to explain.  There was a PRR train during WW-2 that derailed going northbound at this location with many injuries.  How could Amtrak not think that a new loco which could accelerate faster than a GG-1 not also derail there and needed ATC protection ?

 

 

 

 

If an engineer forgets where he is, he might fail to slow down for the curve.

The system designers reasoned that the automatic speed protection system was needed for the southbound route because trains would be approaching at the highest speed limit and the engineer might forget to slow down.

For the northbound route, the highest approaching speed limit was 80 mph, and if a train failed to slow down, it could still successfully negotiate the curve at 80 mph; so they reasoned that the system was needed.

Therefore, while protecting the southbound route from an engineer forgetting where he was, and failing to slow down for the curve; they did not protect the northbound route from an engineer forgetting where he was and accelerating to the highest speed limit. 

 

 

 

 

Makes Amtrak somewhat culpable for not protecting northbound moves as well. They knew there could be a problem and didn't prevent it.

 

I think it was at least a mistake to overlook the fact that the reasoning they applied to protecting the southbound route also applied to the northbound route; perhaps equal to the mistake the engineer made if he forgot where he was. 

[Norm48327]

Euclid

 

blue streak 1

There is something Amtrak needs to explain.  There was a PRR train during WW-2 that derailed going northbound at this location with many injuries.  How could Amtrak not think that a new loco which could accelerate faster than a GG-1 not also derail there and needed ATC protection ?

 

 

 

 

If an engineer forgets where he is, he might fail to slow down for the curve.

The system designers reasoned that the automatic speed protection system was needed for the southbound route because trains would be approaching at the highest speed limit and the engineer might forget to slow down.

For the northbound route, the highest approaching speed limit was 80 mph, and if a train failed to slow down, it could still successfully negotiate the curve at 80 mph; so they reasoned that the system was needed.

Therefore, while protecting the southbound route from an engineer forgetting where he was, and failing to slow down for the curve; they did not protect the northbound route from an engineer forgetting where he was and accelerating to the highest speed limit. 

 

Makes Amtrak somewhat culpable for not protecting northbound moves as well. They knew there could be a problem and didn't prevent it.

[Euclid]

blue streak 1

There is something Amtrak needs to explain.  There was a PRR train during WW-2 that derailed going northbound at this location with many injuries.  How could Amtrak not think that a new loco which could accelerate faster than a GG-1 not also derail there and needed ATC protection ?

 

 

If an engineer forgets where he is, he might fail to slow down for the curve.

The system designers reasoned that the automatic speed protection system was needed for the southbound route because trains would be approaching at the highest speed limit and the engineer might forget to slow down.

For the northbound route, the highest approaching speed limit was 80 mph, and if a train failed to slow down, it could still successfully negotiate the curve at 80 mph; so they reasoned that the system was needed.

Therefore, while protecting the southbound route from an engineer forgetting where he was, and failing to slow down for the curve; they did not protect the northbound route from an engineer forgetting where he was and accelerating to the highest speed limit. 

[Paul_D_North_Jr]

Congressional Limited in 1943 - cause was a burned-off journal a/k/a "hot box".  Speed of 56 MPH had nothing to do with the cause of the derailment (though it sure increased the toll after the inital event occurred - 79 killed, 10 times as many as in the Amtrak wreck): https://en.wikipedia.org/wiki/1943_Frankford_Junction_train_wreck 

- Paul North.