Trains.com

Amtrak 501 Derail in Washington State

74695 views
1887 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    May 2003
  • From: US
  • 25,279 posts
Posted by BaltACD on Tuesday, December 19, 2017 8:25 AM

Saturnalia
For all the evidence pointing to a simple overspeed disaster, there is no evidence that the engineering of the route had anything to do with the incident. 

As I said in the comments of the NewsWire article, I'd like to know how soon PTC was due to be installed. For all of the upgrades going on over this line, why was PTC not turned on before revenue service begins? Not that it is a requirement until the end of next year, and we can't place fault in it not being there, but it'd be an interesting piece to the puzzle.

Like Amtrak 188 on the NEC, it appears as though both curves were due to have PTC online and functioning within months after the respective disasters. 

I used to be a PTC skeptic but with three passenger overspeed incidents in as many years, it is clear that the system is needed to protect passenger movements. Consider me still a skeptic on freight-only lines, but when it comes to passenger safety, these are happening too often. 

While we used to anticipate airline disasters on a fairly regular, grim interval, the last fatality-causing Commerical Airline disaster in the US on  a US carrier was all the way back in 2009. It's time to finish patching up what safety holes remain in Rail. Our industry used to be the safest, but now it is not. 

Transportation will always involve risk. You're more likely to drive on your way to the station and trains are extremely safe. But that does NOT excuse us from using the tools and technologies available to us to prevent accidents such as this. 

Before I retired, CSX had activated PTC on several sub divisions that handled Amtrak.  My understanding, at that time, was that Amtrak locomotives either not had PTC equipment installed or if installed it was not activated as the Amtrak crews had not been trained on it's use; therefore Amtrak's were not operating under PTC.

I have no idea what the PTC situation is/was on the area that the Sounders operate.  PTC has never been something where you just go to Wal-Mart and grab a plug-n-play device from the electronics department, plug it into the engines computer and go play. 

Never too old to have a happy childhood!

              

  • Member since
    May 2003
  • From: US
  • 2,593 posts
Posted by PNWRMNM on Tuesday, December 19, 2017 9:03 AM

All,

I find I have to correct myself. WSDOT says rail cars were in southbound lane of I-5. That makes the engine we see the REAR unit not the lead as I previously thought.

That makes the dynamics of the Ax pretty straight forward. Lead unit tipped over taking the head portion of the train with it.

Sorry for the error in reading the location!

Mac

  • Member since
    September 2010
  • 2,515 posts
Posted by Electroliner 1935 on Tuesday, December 19, 2017 9:23 AM

PNWRMNM
Lead unit tipped over taking the head portion of the train with it.

What I woke up this morning, I could not get out of my head, the question of why the Sieman's locomotive's trajectory shows no deflection from a straight line from the start of the curve. In the Philadelphia wreck, if I remember correctly, I thought the engine left the rails at a point after the beginning of the curve. Shouldn't there have been some wheel/rail force to start the locomotive to begin turning before climbing the rail? I know it's momentum wants it to go in a straight line and this looks like it did but I don't understand why it didn't start the curve before climbing up over it. Physics anyone. 

  • Member since
    July 2008
  • 2,325 posts
Posted by rdamon on Tuesday, December 19, 2017 9:32 AM

If you look at the line (tangent) the curve appears to start further back ..

 

 

  • Member since
    January 2014
  • 8,217 posts
Posted by Euclid on Tuesday, December 19, 2017 9:58 AM

What line of tangent are you referring to?  How can these photos show the path taken by the derailed locomotive?  I assume the derailment happend about half way through the curve

  • Member since
    December 2007
  • From: Georgia USA SW of Atlanta
  • 11,919 posts
Posted by blue streak 1 on Tuesday, December 19, 2017 9:59 AM

have a question about the speed limit signs.  The one picture we see looking south toward the p-42 shows the speed sign as T-30 ; P-30.  That sign appears only at most 1000 feet from start of curve and maybe less.   Were there additional signs further north showing slower speeds ?  1000 feet is not enough to go from 80 MPH to 30 MPH ?

Anyone know what the BNSF requirements for a slower speed posting in distance from the slower speed ?

Use BNSF since they apparently  dispatch the Lakewood spur.

  • Member since
    December 2006
  • 1,754 posts
Posted by diningcar on Tuesday, December 19, 2017 10:01 AM

 rdamon, if the photo is correctI Agree. At the location of the RR bridge over southbound I-5 the train is exiting the curve with tangent track going south toward the I-5 northbound bridge.

  • Member since
    July 2008
  • 2,325 posts
Posted by rdamon on Tuesday, December 19, 2017 10:17 AM

They did quite a bit of work on the track . as compared to this 2012 street view

 

 

 

  • Member since
    April 2003
  • 305,205 posts
Posted by Anonymous on Tuesday, December 19, 2017 10:19 AM

Euclid
What line of tangent are you referring to? How can these photos show the path taken by the derailed locomotive? I assume the derailment happend about half way through the curve

In the video linked before at about 20 - 30 seconds show a good aerial view of the curve and the path the SC44 took: https://youtu.be/MZVaRUBceLg

With 2.6 times the allowed speed it is possible the train derailed already within the transition curve. And at a 8° curve that is quite short: L = 1.63 x Eu x V  Eu in Inches, V in mph, L in feet

IIRC the unbalanced elevation can be 3'' to 9'' for passenger trains. How much it was here I don't know.
Regards, Volker

  • Member since
    July 2006
  • 20 posts
Posted by MOWBill on Tuesday, December 19, 2017 10:27 AM

2 miles in advance, sign is diagonal.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, December 19, 2017 10:59 AM

PNWRMNM
That makes the dynamics of the Ax pretty straight forward. Lead unit tipped over taking the head portion of the train with it.

Mac, I think you need some more correction.  The SC44 arrived at the bottom of its trip upright and facing forward, with no particular sign of rollover damage.  So I think 'tipping over' isn't the dominant derailment physics here; I have not looked carefully at the right-side panels or the undercarriage of the unit to see if it shows either dirt corresponding to the 'divots' roughly adjacent to the cabin, or damage corresponding to wiping out the signal bridge.  The path the unit took was likely partially constrained by the berm and other topography, and as near as I can tell the first several cars followed it in a straight line (which to me indicates it had substantial momentum and the drawbar and intercar articulation connections involved remained under tension 'all the way into the woods' where we see them.

I do suspect that the dead mass of the P42 acted to push the rear of the consist into the jackknifing we see, and it will be interesting to see the recreation of the forces in the accident when it is developed.  Interestingly the back end came to a stop 'on the rails' with at least one car preceding it, which reflects a short stop indeed were braking from the supposed 81mph only initiated close to the curve.

I have to wonder if some of the speculation about 'an impediment on the track' refers to the very early departure of the SC44 from the track, if indeed the locomotive was the part of the train that took out the signal bridge.  I see little if no way that all the jackknifed parts of the train could wind up how and where they are if they had previously 'excursed' all the way over to that point of contact...

  • Member since
    January 2014
  • 8,217 posts
Posted by Euclid on Tuesday, December 19, 2017 11:02 AM

MOWBill

2 miles in advance, sign is diagonal.

 

What does that sign say?  My point is that if the speed limit goes from 79 to 30, a train must begin reducing speed before it enters that 30 mph zone.  I assume that an advance warning sign would say something that tells the engineer to begin slowing down to reach a speed not exceeding 30 mph at some distance ahead.

What I would like to know is how fast the train was going when it entered the curve.  We know that it was going 80 in a 30 zone.  So I would like to know exactly where that 30 zone begins in relation to the start of the curve.

If the 30 zone begins right at the curve, then I know that the train entered the curve at 80.  However if the 30 zone begins at 1000 feet from the curve, then I only know that the train was going 80 when 1000 feet from the curve.  From that point, the engineer may have braked and entered the curve at maybe 55-65 for instance.

  • Member since
    October 2016
  • 185 posts
Posted by Saturnalia on Tuesday, December 19, 2017 11:03 AM

Euclid

What line of tangent are you referring to?  How can these photos show the path taken by the derailed locomotive?  I assume the derailment happend about half way through the curve

 

So what you're going to find is that trains derailing like this generally won't make it halfway through the curve, particularly where the curve in question changes heading signfiicantly. That's because once they're into the full force of the curve, it won't take long for inertia to do the dirty work. 

For those unaware, railroad curves are not like curves on the freeway which are literally just arcs of a circle. Railroad curves utilize a "spiral" to transition from straight (tangent) track to the full radius of the curve. This spiral is also used to transition the superelevation. The design of this spiral, and indeed the rest of the curve, is generally laid out according to the AREMA standard, which is the standard the whole industry uses here in the US and I'd wager a significant portion of the rest of the world. These design tables also include maxiumum authorized speed for those curves. This is a time-tested series of rules which has done a great job keeping trains on the track! 

I hate to say it, but with the NTSB confirming what most of us already picked out - that the train went 80 into a 30 mph curve, it is becoming clear that engineering of the area was not to blame, and unless they find that the train's brakes somehow failed, it's looking to be another case where the engineer did not, for whatever reason, follow the established timetable and wayside signage. This will be the third in recent years, since the Metro North and Amtrak 188 disasters. All three PTC-preventable, all three occured just months before PTC was due to be activated. 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, December 19, 2017 11:04 AM

Electroliner 1935
What I woke up this morning, I could not get out of my head, the question of why the Siemens locomotive's trajectory shows no deflection from a straight line from the start of the curve. In the Philadelphia wreck, if I remember correctly, I thought the engine left the rails at a point after the beginning of the curve. Shouldn't there have been some wheel/rail force to start the locomotive to begin turning before climbing the rail? I know it's momentum wants it to go in a straight line and this looks like it did but I don't understand why it didn't start the curve before climbing up over it. Physics anyone.

I am beginning to wonder about this, too, with the additional detail that at the Philadelphia wreck, if I remember correctly, the brake was not applied until the train was substantially into the curve.  Could an excessive application of blended brake coupled with weight transfer on the truck result in a compromise of axle dynamics leading to a combination of early flange lift and the 'steering' of the sideframe relative to the carbody popping the leading flange over the railhead?  Would likely only happen in emergency braking combined with compromised adhesion, but could be 'tested for'.

  • Member since
    January 2014
  • 8,217 posts
Posted by Euclid on Tuesday, December 19, 2017 11:09 AM

Saturnalia
I hate to say it, but with the NTSB confirming what most of us already picked out - that the train went 80 into a 30 mph curve,...

The NTSB has not confirmed that.  My question above is related to this.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, December 19, 2017 11:24 AM

Euclid
The NTSB has not confirmed that.

OSAF:

 

Euclid
My question above is related to this.

Didn't 'MOW Bill' already reply to this: 2 mile advance with a slant board?  That is the initial warning a restriction is coming up; the 'actual notice' is the board right ahead of the restriction ... whether you like that, or not (as happens, I don't).

I don't think this has much to do with whether or not the train was overspeeding at the point of the T30 P30 marker, however; that would be have to be reconstructed from the observed incident physics to a degree, but it would take substantial momentum to get the SC44 where it wound up, with the first several cars of the lightweight consist still following it 'straight' into the woods instead of starting to jackknife or tumble.

I agree that we should wait for the NTSB preliminary report, as I think they will have both the GPS tracking data and the information from the LER* to tell us EXACTLY how fast the train was going at all the points of interest, what the brake transitions were, and most of the other information related to the question you thought you were asking.

 

*I am sure some clever no-sh*t-Sherlock will be claiming "the locomotive went off the rails so the data won't apply".  The LER in question will be the one in the trailing locomotive, which did NOT go off the rails.

  • Member since
    July 2006
  • From: North Dakota
  • 9,592 posts
Posted by BroadwayLion on Tuesday, December 19, 2017 11:25 AM

PNWRMNM
The engine on the track was the LEAD unit. Since there are some Talgo cars off the high side of the curve North of the freeway and the rear unit looks to be south of the freeway,

 

This was a south bound train, the LEAAD unit is on the freeway sitting on top of a tractor-trailer unit. Cars followed him into the wild blue yonder. The trailing locomotive was on the REAR of the train, and helped push the rear end of the train into an accordion.

 

The message from the engineer was "Emergency, Emergency, Emergency, We are on the ground"

He did not say anything about flying through the woods or landing on top of a Freightliner. He said he was "On The Ground"

From the time the train reache the curve until it landed on the interstate there would not have been time to give any sort of an emergency broadcast.

Once off the rails (wherever that may have happened) neither the speed recorder nor the brakes would have had effect.

 

Thus says the LION, your truths may vary.

ROAR

The Route of the Broadway Lion The Largest Subway Layout in North Dakota.

Here there be cats.                                LIONS with CAMERAS

  • Member since
    September 2017
  • 5,636 posts
Posted by charlie hebdo on Tuesday, December 19, 2017 11:26 AM
Exactly! Whatever other factors (human error, equipment malfunction, etc.) might be in play, the keys are 80mph in 30mph curve; no PTC.
  • Member since
    May 2003
  • From: US
  • 25,279 posts
Posted by BaltACD on Tuesday, December 19, 2017 11:30 AM

Euclid
If the 30 zone begins right at the curve, then I know that the train entered the curve at 80.  However if the 30 zone begins at 1000 feet from the curve, then I only know that the train was going 80 when 1000 feet from the curve.  From that point, the engineer may have braked and entered the curve at maybe 55-65 for instance.

View the Rochelle Web Cam - the two Yellow signs on the BNSF tracks indicate the starting point for the Speed Restriction over the diamond.

Permanent Speed Restriction signs are placed AT the point the restriction begins.  There may be advance warning signs of the upcoming speed restriction and there may not be advance warning signs - each carrier has its own standards.

Never too old to have a happy childhood!

              

  • Member since
    January 2014
  • 8,217 posts
Posted by Euclid on Tuesday, December 19, 2017 11:44 AM

Overmod
I agree that we should wait for the NTSB preliminary report, as I think they will have both the GPS tracking data and the information from the LER to tell us EXACTLY how fast the train was going at all the points of interest, what the brake transitions were, and most of the other information related to the question you thought you were asking.

I am quite certain about what question I am asking. It is a very simple question.  It has not been answsered as far as I know.  And I don't intend to wait for the NTSB to answer this incredibly simple question.  Here is that question:

At what point does the speed limit change from 79 to 30?

  • Member since
    February 2002
  • From: Mpls/St.Paul
  • 13,892 posts
Posted by wjstix on Tuesday, December 19, 2017 11:52 AM

x1217

VOLKER LANDWEHR
 
Euclid
What line of tangent are you referring to? How can these photos show the path taken by the derailed locomotive? I assume the derailment happend about half way through the curve

 

In the video linked before at about 20 - 30 seconds show a good aerial view of the curve and the path the SC44 took: https://youtu.be/MZVaRUBceLg

With 2.6 times the allowed speed it is possible the train derailed already within the transition curve. And at a 8° curve that is quite short: L = 1.63 x Eu x V  Eu in Inches, V in mph, L in feet

IIRC the unbalanced elevation can be 3'' to 9'' for passenger trains. How much it was here I don't know.
Regards, Volker

 

The video at 5:40 shows the roof of the lead engine basically entirely ripped off, and at 8:45 you can see damage to the right side of the lead engine. At 7:50 you can see one or more trees knocked over, with one tree having a huge gouge taken out of it about 10 feet up from the base. I'd speculate the engine could have tilted to the right when it left the tracks, became airborne, hit the trees (causing the roof to be scraped off) but then - either due to hitting the trees or the slope of the hill leading down to the road beneath the bridge - ended up back on it's wheels.

Is there any chance the problem could have been with the rear engine - the one still on the rails? Like the engineer in the lead unit throttling down and applying brakes when he should have, but somehow the rear engine not responding due to some mechanical failure.

Stix
  • Member since
    December 2001
  • From: Denver / La Junta
  • 10,820 posts
Posted by mudchicken on Tuesday, December 19, 2017 12:08 PM

...There's this thing called an employee timetable ... the locomotive operator/ engineer was supposed to be intimately familiar with it. The event recorder will tell quite a story, including if there was even the beginning of a brake application or throttle reduction before centripital force took over.

Volker: 3 inch unbalance. (FRA rule 213.57 at up to Class 5 (this was a Cls 2 curve in otherwise Cls 4 territory, Eu doesen't start to change until you hit Cls 6 and only then with restrictions per FRA), freight rail normally is more restrictive and less forgiving at 1 3/4" to 2  1/4"; BNSF is at 2", but this was not their railroad anymore) ... Documentation for the FRA allowing something other than 3" unbalance in the TALGO equipment with get a hard look by NTSB.

I'll bet BNSF is going to go over its vetting process on allowing foreign rail operating people over its road. There have been little issues before in other places. Who else also should have been in the cab on that first revenue run will also get some scrutiny.

The reports of people saying that the trainset hit something was probably the derailed train striking the butt end of the thru plate deck girder bridge.

Mudchicken Nothing is worth taking the risk of losing a life over. Come home tonight in the same condition that you left home this morning in. Safety begins with ME.... cinscocom-west
  • Member since
    December 2001
  • From: Denver / La Junta
  • 10,820 posts
Posted by mudchicken on Tuesday, December 19, 2017 12:10 PM

Euclid
 
Overmod
I agree that we should wait for the NTSB preliminary report, as I think they will have both the GPS tracking data and the information from the LER to tell us EXACTLY how fast the train was going at all the points of interest, what the brake transitions were, and most of the other information related to the question you thought you were asking.

 

I am quite certain about what question I am asking. It is a very simple question.  It has not been answsered as far as I know.  And I don't intend to wait for the NTSB to answer this incredibly simple question.  Here is that question:

At what point does the speed limit change from 79 to 30?

 

Read Balt's post again. Carefully this time.

Mudchicken Nothing is worth taking the risk of losing a life over. Come home tonight in the same condition that you left home this morning in. Safety begins with ME.... cinscocom-west
  • Member since
    June 2007
  • 228 posts
Posted by RDG467 on Tuesday, December 19, 2017 12:16 PM

Overmod
 
PNWRMNM
That makes the dynamics of the Ax pretty straight forward. Lead unit tipped over taking the head portion of the train with it.

 

Mac, I think you need some more correction.  The SC44 arrived at the bottom of its trip upright and facing forward, with no particular sign of rollover damage.  So I think 'tipping over' isn't the dominant derailment physics here; I have not looked carefully at the right-side panels or the undercarriage of the unit to see if it shows either dirt corresponding to the 'divots' roughly adjacent to the cabin, or damage corresponding to wiping out the signal bridge.  The path the unit took was likely partially constrained by the berm and other topography, and as near as I can tell the first several cars followed it in a straight line (which to me indicates it had substantial momentum and the drawbar and intercar articulation connections involved remained under tension 'all the way into the woods' where we see them.

I do suspect that the dead mass of the P42 acted to push the rear of the consist into the jackknifing we see, and it will be interesting to see the recreation of the forces in the accident when it is developed.  Interestingly the back end came to a stop 'on the rails' with at least one car preceding it, which reflects a short stop indeed were braking from the supposed 81mph only initiated close to the curve.

I have to wonder if some of the speculation about 'an impediment on the track' refers to the very early departure of the SC44 from the track, if indeed the locomotive was the part of the train that took out the signal bridge.  I see little if no way that all the jackknifed parts of the train could wind up how and where they are if they had previously 'excursed' all the way over to that point of contact...

 

Overmod, I think that's a good analysis based on the photos provided. I'd add that the Talgo set 'buckled' as the SC44 came to a sudden stop (see the naked wheel tilt mechanism at the bridge girder) and the interia of the second part of the train+P42 shoved the rest of the set off the left side of the bridge. I think the upside-down carriage was the one which was shorn of its WTM.

I thought I saw some track damage before the curve, but that could've been a photographic illusion.

  • Member since
    January 2014
  • 8,217 posts
Posted by Euclid on Tuesday, December 19, 2017 12:24 PM

mudchicken
  Read Balt's post again. Carefully this time.

Okay, the 30 mph limit begins right at the curve, so we know the train entered the curve at 80 mph.  So the wreckage is the result of an 80 mph derailment.  No braking slowed the train before reaching the curve.

  • Member since
    September 2010
  • 2,515 posts
Posted by Electroliner 1935 on Tuesday, December 19, 2017 12:30 PM

The tangent I am referring to is to the center line of track North East of the curve. It runs through the destroyed signal and past the signal hut and into the end of the cars. It supprises me that there appears that the Locomotive did not appear to have any deflection to its left but stayed straight ahead. Does not make sense to me.My recollection of the Philadelphia wreck was that the engine started the curve before it left the rails and it was going faster than 80. While the Charger is heavier (132 T vs 108 T) than the ACS, I don't think that can account for this. Does anyone else think there should have been some leftward deflection from the tangent IF the front truck was on the rail? This concerns me. 

  • Member since
    July 2008
  • 2,325 posts
Posted by rdamon on Tuesday, December 19, 2017 12:31 PM

This looks like the advance signs

 

 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, December 19, 2017 12:45 PM

mudchicken
Read Balt's post again. Carefully this time.

In all fairness, that didn't answer the question definitively, just framed it properly.  What Euclid needs is the precise type and distance of 'advance warning' for that section of track as provided by carrier BNSF.

What I suspect the 'exact' answer to that is going to be has, I think, already been provided, but since I don't know for sure I will lay it out for correction by those who do:

1) Speed restrictions on that line apply to the immediately following section (i.e. not including any 'deceleration range') -- they do not correspond to a distant warning but a 'home' signal.

2) As noted, permanent restrictions are part of an engineer's training, and one of the reasons why 'becoming qualified on the territory' (and the months-long training of crews up to the service inauguration) is so important.  This is not a MUTCD situation where any old person speeding up on a hazard "has" to be warned to check speed in advance; the sign marks a known hazard and reminds what the known speed through it is to be.

3) According to the earlier reference, there is an indication, presumably that a 'fixed restriction' is upcoming, at some distance ahead of the curve.  I doubt that this will have physical speeds for Talgo and Passenger indicated on it, however; it would be a reminder that something that should be known is coming up.

Now, a PTC enablement would of necessity have some programmed deceleration length for the SERVICE application reducing from 79mph to 30mph, and the setpoint for either ATC or ATS would be no less than that distance ahead of the permanent notification.  Any fixed 'reduce to' type of speed notification would probably factor in some reasonable reaction time (at a nominal ~100fps if you want it in distance terms).

There is a disparity in high-speed notification, too: there are curve restrictions near New Brunswick in the NEC that are like the famous 'slow to 90' boards on MILW -- the difference being that the NEC markers are directly ahead of the curve, with little more 'advance warning' given the speed differential than at the Lakewood curve.

I suspect that what you want the answer to your question to be is different from the objective 'answer' that the rules provide.  It would appear that the 'speed limit' of 30 mph applies only to the hazard, and the crew 'training' is what establishes how and where 'way' is taken off a train moving 79mph to get it going no faster than 30mph when it encounters the hazard.  Now we need to see the 'other side' of the signs rdamon has found to see if the 'slant board' warning does in fact carry speed or reduction information adjacent to it.

  • Member since
    September 2010
  • 2,515 posts
Posted by Electroliner 1935 on Tuesday, December 19, 2017 1:00 PM

When yo drive on a highway, you see warning signs for XX SPEED ZONE AHEAD, then 500 to a 1000 feet further, you see the SPEED LIMIT XX, The vehicle should pass the SPEED LIMIT sign at XX speed. Same for a train. Is that clear? 

  • Member since
    January 2014
  • 8,217 posts
Posted by Euclid on Tuesday, December 19, 2017 1:36 PM

The reason I asked where the 30 limit began is because we know the train was running 80 at that point.  So I wondered if the engineer had any time to apply braking sufficient to begin slowing the train prior to the derailment. 

I can see losing awareness of the impending curve even after passing the advance warning.  But in full daylight, with straight track leading to that abrupt curve, there must be at least 1000 feet of visual indication of the curve ahead.  I would think it most probable that an engineer would realize what was happening and would apply maximum braking at least 500 feet ahead of the curve, if not further.

But if he was going 80 when he reached the 30 zone, he was apparently completely unaware of what was happening until reaching the point of derailment.  That seems highly improbable.     

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy