Amtrak Derailment Philadelphia (5/12/2015)

[Dutchrailnut]

yes except the alerter would have put train in penalty brake.

 

[WoodyinNYC]

NTSB Issues Second Update on its Investigation Into the Amtrak Derailment in Philadelphia6/10/2015

WASHINGTON – As part of its ongoing investigation into the May 12, 2015, derailment of Amtrak Train 188 in Philadelphia, the NTSB today provides this update on the analysis of the engineer’s cell phone and related records.

The NTSB is conducting a detailed examination of the engineer’s cell phone calls, texts, data and cell phone tower transmission activity records from the phone carrier; and records from Amtrak’s on-board Wi-Fi system.

Analysis of the phone records does not indicate that any calls, texts, or data usage occurred during the time the engineer was operating the train. Amtrak’s records confirm that the engineer did not access the train’s Wi-Fi system while he was operating the locomotive.

...

-----------------------

http://www.ntsb.gov/news/press-releases/Pages/PR20150610.aspx

Boldface added by me, because I feel real good about this news.

 

[jis]

That is good to be reconfirmed. That is something that I had suspected all along. The focus was on it simply because one of the recent accidents involved texting. I bet they also carefully verified that he did not fall asleep too!

I bet the Philadelphia Police Department is disappointed. Afterall that was one of their pet theories to justify their outbursts about "reckless driving".

 

[crescent2]

Woody, I do, too.

http://news.yahoo.com/ntsb-says-amtrak-engineer-didnt-132215729.html#

On another note, how often does the alerter require a response? If mechanical issues could be ruled out, I have a hard time believing the engineer wasn't incapacitated in some way. Why the train accelerated in the last minute is puzzling.

 

[BCL]

NTSB Issues Second Update on its Investigation Into the Amtrak Derailment in Philadelphia

6/10/2015

WASHINGTON – As part of its ongoing investigation into the May 12, 2015, derailment of Amtrak Train 188 in Philadelphia, the NTSB today provides this update on the analysis of the engineer’s cell phone and related records.

The NTSB is conducting a detailed examination of the engineer’s cell phone calls, texts, data and cell phone tower transmission activity records from the phone carrier; and records from Amtrak’s on-board Wi-Fi system.

Analysis of the phone records does not indicate that any calls, texts, or data usage occurred during the time the engineer was operating the train. Amtrak’s records confirm that the engineer did not access the train’s Wi-Fi system while he was operating the locomotive.

...

-----------------------

http://www.ntsb.gov/news/press-releases/Pages/PR20150610.aspx

Boldface added by me, because I feel real good about this news.

The engineer claimed that he had his personal phone stashed in his backpack, as required by Amtrak rules. I'm thinking a lot of passenger train systems got tough on that after one of those Metrolink collisions where it was revealed that an engineer would text on the job to rail fans.

Also, the only photo they show of him was as a conductor. I've seen several Amtrak engineers getting ready to relieve someone or getting out of a cab, and the typical "uniform" is Silicon Valley like - most often a T shirt and jeans. Maybe a baseball style cap.

 

[AlanB]

On another note, how often does the alerter require a response? If mechanical issues could be ruled out, I have a hard time believing the engineer wasn't incapacitated in some way. Why the train accelerated in the last minute is puzzling.

Since the throttle was being advanced approaching the curve, based upon the evidence that the train was speeding up, that input to the control systems would have satisfied the alterter and would have continuously reset the timer.

However to answer your question, while I believe a RR can program what they like, generally the alterer waits a random amount of time between 1 minute and 2 minutes before it goes off. So it could wait 1:20 this time, 1:40 next time, roll back to 1:25, etc. And of course as already mentioned, any action by the engineer such as blowing the horn, ringing the bell, braking, adjusting the throttle up or down, will reset the timer without the alterer's alarm going off. If no actions are taken, then the engineer must hit a button to satisfy the alterer when the timer runs out.

 

[crescent2]

Thanks, Alan.

So, if the engineer had some medical issue and fell forward onto the throttle, that could account for the acceleration just before entering the curve?

(I want this engineer to not be at fault!)

 

[WoodyinNYC]

Woody, I do, too.

...

If mechanical issues could be ruled out, I have a hard time believing the engineer wasn't incapacitated in some way. Why the train accelerated in the last minute is puzzling.

I can imagine that the engineer made some error that contributed to the crash. We await more information from the Safety Board.

But at least he didn't make the cell-phone-use error that would have been a BIG, stupid, carelessly evil mistake.

 

[AlanB]

So, if the engineer had some medical issue and fell forward onto the throttle, that could account for the acceleration just before entering the curve?

(I want this engineer to not be at fault!)

First, I'm not sure if that's even possible. I'm not an engineer, but I think that it takes more than just a forward thrust to move the throttle. I know on many light rail train one has to push in a button first to release the control before it will move.

Second, even if it were possible, you would see a sudden quick burst of speed and that didn't happen here. It seemed like a more steady, measured speed increase. Almost like the engineer forgot where he was on the line and thought he was past the curve and needed to accelerate to the higher track speed after the curve. Please note that is pure speculation on my part. I'm not trying to blame the engineer either.

Finally, had he fallen against it and been able to shove it all the way forward, that would have ended his input to the controls. So the alterer timer would have been running again and most likely gone off before the curve.

 

[FormerOBS]

There's a remote possibility that I'm wrong about some aspects of my assessment, but I'm pretty certain this is essentially what happened:

The engineer wouldn't have to fall against the throttle to accelerate. As I understand it, after leaving 30th St. Station, the engineer typically advances the throttle so that the train gradually accelerates to around 75 - 80 m.p.h., then he eases back on the throttle to reduce speed for the 50 m.p.h. curve. If he is injured or distracted during the acceleration phase, he might neglect to reduce the speed after reaching 75 - 80 m.p.h., and the engine will continue to increase speed until the engineer initiates a brake application, or until the alerter does it for him. From all indications, it appears that he did not reduce the throttle setting after reaching 75 - 80 m.p.h. The alerter did not sound, probably because this all happened in too short a time span. It would have come into play a short time later.

The question is not so much what happened, as why.

Tom

 

[George K]

If I recall, NTSB said that there was no mechanical failure of the equipment.

What other explanations are possible for the events?

 

[SarahZ]

There's a remote possibility that I'm wrong about some aspects of my assessment, but I'm pretty certain this is essentially what happened:

The engineer wouldn't have to fall against the throttle to accelerate. As I understand it, after leaving 30th St. Station, the engineer typically advances the throttle so that the train gradually accelerates to around 75 - 80 m.p.h., then he eases back on the throttle to reduce speed for the 50 m.p.h. curve. If he is injured or distracted during the acceleration phase, he might neglect to reduce the speed after reaching 75 - 80 m.p.h., and the engine will continue to increase speed until the engineer initiates a brake application, or until the alerter does it for him. From all indications, it appears that he did not reduce the throttle setting after reaching 75 - 80 m.p.h. The alerter did not sound, probably because this all happened in too short a time span. It would have come into play a short time later.

The question is not so much what happened, as why.

Tom

This is what I've been thinking all along. He gradually increased the speed, [something] happened to prevent him from slowing the train down, and then [something] stopped happening and he hurried to slow/stop the train, knowing that curve was approaching.

 

[crescent2]

Yes, my feeling (hope?) is that something incapacitated him for a short time, causing the train to accelerate too much before the curve.

 

[Just-Thinking-51]

Or he got up to see the damage to his fireman window from the rock, while the train was power up in speed. Talking on the radio would meet the alerter needs. Got back to his seat saw the curve hit the emergency brakes and derailed the train.

Only need a few seconds of inattention....

If you read his Trainorders post he has quite aware of the dangers and how easy it was to screw up. He was quite critical of the lack of PTC. Old or New system.

 

[AlanB]

Without going to look at what the timeline showed, and the NTSB hasn't provided the entire timeline yet, I got the impression that he was cruising along at the indicated 80 MPH track speed for a bit. Then the speed was increased starting a minute or so before the crash. That would tend to discount the idea that he moved the throttle forward after crossing the Schuylkill River and entering the straightaway that leads to the curve and just left it in such a position so as to allow the speed to continuously increase.

Additionally, depending on what notch one puts the throttle in, there is a limit to just how fast the train will go. I can't imagine that he would put the throttle into the top notch that would bring the engine to 125 MPH knowing that the top speed was 80 MPH. He'd select the notch setting that gets him closest to 80 MPH.

Unlike a subway train where you slam the throttle to maximum to accelerate as fast as possible and the slam on the brakes for the next station, Amtrak engineers tend to accelerate more smoothly and with a few less G's. They don't slam the throttle to the top notch as they leave a station; they ramp up to provide a more comfortable ride.

 

[PRR 60]

Without going to look at what the timeline showed, and the NTSB hasn't provided the entire timeline yet, I got the impression that he was cruising along at the indicated 80 MPH track speed for a bit. Then the speed was increased starting a minute or so before the crash. That would tend to discount the idea that he moved the throttle forward after crossing the Schuylkill River and entering the straightaway that leads to the curve and just left it in such a position so as to allow the speed to continuously increase.

Additionally, depending on what notch one puts the throttle in, there is a limit to just how fast the train will go. I can't imagine that he would put the throttle into the top notch that would bring the engine to 125 MPH knowing that the top speed was 80 MPH. He'd select the notch setting that gets him closest to 80 MPH.

Unlike a subway train where you slam the throttle to maximum to accelerate as fast as possible and the slam on the brakes for the next station, Amtrak engineers tend to accelerate more smoothly and with a few less G's. They don't slam the throttle to the top notch as they leave a station; they ramp up to provide a more comfortable ride.

The line from the Schuylkill River to Frankford Junction is not really a straightaway. There are multiple speed restrictions in that section prior to getting to the accident location.

From the Schuylkill River to roughly North Philadelphia Station (PHN) the maximum track speed is 80mph. Through PHN, the track speed drops to 60. Then it's back up to 80 for about a mile, then down to to 65 for the MP84 (Second Street) curve, then back up to 80 for about 2-1/2 miles until the 50mph curves at Frankford Junction. The data recorder will be able to show how the train was handled through the various speed restrictions leading to Frankford junction.

 

[AlanB]

Agreed Bill, but it is more of a straightaway than the first quarter of a mile or so to the river after leaving PHL and prior to the curve. Guess I should have been clearer.

 

[PRR 60]

Agreed Bill, but it is more of a straightaway than the first quarter of a mile or so to the river after leaving PHL and prior to the curve. Guess I should have been clearer.

My point was that the train had several speed-control requirements between the river and Frankford Junction. It is not simply setting the speed to 80 and sitting there for six miles. The train should have gone through several accelerations and decelerations prior to getting to the accident scene. Adherence to those speed requirements will be shown by the data recorder.

One correction to my prior post: from the river to PHN, the maximum track speed is 80mph for Acela, but 70mph for Regionals. Here's how it breaks down for the Regional from the river to east of Frankford Junction:

- MP87.2 Mantua (at the Schuylkill River) to MP85.1 Lehigh (PHN) - Track speed 70mph

- MP85.1 Lehigh to MP 84.5 Clearfield - Track speed 60mph

- MP84.5 to MP82.1 Shore (just west of Frankford Junction) - Track speed 80mph

*** Curve at MP84 - 65mph

- MP82.1 Shore to MP77.2 Holmes - Track speed 110mph

*** Curve at Shore - 50mph <== accident curve

*** Curve at MP 81 - 60mph <== just past accident curve

*** Speed restriction below established track speed.

My theory is that the engineer lost track of where he was on the railroad. He could have been distracted by anything, including his own thoughts. It was night, and the area lacks nighttime visual references. The MP84 curve is geometrically similar to the MP81 curve. As he came out of the MP84 curve, he may have thought almost subconsciously that he was coming off the MP81 curve and entering 110mph territory. The next restriction after the MP81 curve are 90mph curves six miles ahead. He routinely accelerated to near 110mph (106). Less then two minutes later, the Frankford Junction curve was directly ahead in his headlights. He hit emergency, but it was too late.

 

[jis]

PRR I had been thinking the same for quite a while, and mentioned it very early on in this thread. But people who claimed to know the territory better but quite evidently were quite clueless, dissed the hypothesis, and since in the bigger scheme of things it didn't matter whether these ignorant folks agreed or not, I just let it drop. I actually do that quite often these days. Saves a lot of time.

 

[Mystic River Dragon]

jis--on the other hand, you and PRR 60 had the company of a lot of other intelligent people thinking exactly what you both were thinking. It makes more sense than anything else, unless he had a seizure or something else medical (but the doctors at the hospital should have been able to determine that).

 

[Thirdrail7]

PRR I had been thinking the same for quite a while, and mentioned it very early on in this thread. But people who claimed to know the territory better but quite evidently were quite clueless, dissed the hypothesis, and since in the bigger scheme of things it didn't matter whether these ignorant folks agreed or not, I just let it drop. I actually do that quite often these days. Saves a lot of time.

I do believe people offered other viewpoints. That does not necessarily make them anymore wrong, ignorant or clueless than you, me or anyone else. This is because until we hear from the engineer in question, the various reasons we all are exploring are nothing more than speculation or theories that may happen to fit the facts...as known and/or presented.

 

[niemi24s]

With all those different speed limits in a relatively short distance of about 5½ miles (Post #843) would an engineer jockey the throttle in an attempt to reduce transit time to the bare minimum or simply strike some happy medium using fewer throttle settings? Talking here about what might be considered normal everyday operation and being pretty much on schedule.

 

[Thirdrail7]

With all those different speed limits in a relatively short distance of about 5½ miles (Post #843) would an engineer jockey the throttle in an attempt to reduce transit time to the bare minimum or simply strike some happy medium using fewer throttle settings? Talking here about what might be considered normal everyday operation and being pretty much on schedule.

This is something that I was trying to bring up to Acela150 in post 697 when I mentioned in part:

Where is the normal braking point for the curve? Are you aware that people run trains differently? How much time did that crew have together so the crew could even determine the engineer's style?

I've noticed over the years that running a train isn't a team sport. There are different styles. I know of some people that will squeeze every second out of the schedule, accelerating and decelerating, braking and drawing power between every speed restriction BECAUSE THAT'S THE TRACK SPEED!

I also know people that kind of mosey along, using the the terrain, grades and throttle manipulation to maintain speeds or schedules.

That being said, I know some people that get right up to speed, braking for the restriction through PHN, then accelerate to the point that you can feel them braking for the 65mph curve only to accelerate to track speed before braking for the Frankford curves.

On the other hand, I know people that never operate above 70 mph in the 80 mile mph zone, particularly if they are headed east. I know people that will get up to 70 after Mantua, and will allow gravity to slow the train down for travel through PHN. Once they clear Clearfield Interlocking, they will casually get up to 65 and never touch the throttle until they are on the other side of the Frankford curves and are accelerating towards Holmes.

Why?

One thing that I haven't seen addressed by anyone in this thread (although I may have missed it since this is a long thread) is the grades involved in the territory. It may not look like much, but there is a nice little downhill grade heading east from MP 84 to Shore. So much, that if you're operating west, the 829 and 835 automatics have the "G" (grade) marker so freight trains and/or work trains do not have to actually stop if either of those signals are displaying the stop and proceed aspect. They may proceed as though a restricting aspect is being displayed. Gravity alone will get a train up to around 72mph from the curve so there isn't much loss in just drifting. As such, I know it would take much for a train to reach a great rate of speed in this stretch whether you are throttled out or not. Even an old DC AEM-7 would be able to achieve a good rate of speed.

At the end of the day, it boils down to personal style which may vary from day to day and from train to train. His choice on this particularly will be readily available on the engine download.

 

[jis]

I have read somewhere, in some French railroad journal, that the operators of the TGV Atlantique trains operating into Gare Montparnasse have had an ongoing competition to see who can cut off power the farthest from the bumper post and drift into the station using only brakes as necessary, while meeting all speed restrictions and schedule. You see there is generally a shallow downgrade into Montparnasse from way outside Paris. Since then I have heard various rumors that people have managed to do this from 186mph a little past Rouvray outside Paris, which seems incredible to me, and is testimony to the incredibly low resistance of steel wheel on steel rail. Don't know the truth or falsehood of this, but the whole idea seems quite interesting and a way to test how efficiently one can operate the train energy consumption-wise, while keeping to schedule.

 

[Hal]

I've noticed over the years that running a train isn't a team sport. There are different styles. I know of some people that will squeeze every second out of the schedule, accelerating and decelerating, braking and drawing power between every speed restriction BECAUSE THAT'S THE TRACK SPEED!

I also know people that kind of mosey along, using the the terrain, grades and throttle manipulation to maintain speeds or schedules.

That being said, I know some people that get right up to speed, braking for the restriction through PHN, then accelerate to the point that you can feel them braking for the 65mph curve only to accelerate to track speed before braking for the Frankford curves.

On the other hand, I know people that never operate above 70 mph in the 80 mile mph zone, particularly if they are headed east. I know people that will get up to 70 after Mantua, and will allow gravity to slow the train down for travel through PHN. Once they clear Clearfield Interlocking, they will casually get up to 65 and never touch the throttle until they are on the other side of the Frankford curves and are accelerating towards Holmes.

Why?

The answer to "Why" is experience. It takes years to be experienced. In olden times there were two in the cab. One would have that years of experience and the one without the years would be getting the exerience.