Amtrak derailment in Montana

[hokie66]

Three deaths reported so far.

https://www.foxnews.com/us/amtrak-derailment-montana

 

[FacEngrPE]

NTSB Newsroom, go team sent to site.

https://twitter.com/NTSB_Newsroom/status/1441958985003945986?ref_src=twsrc^tfw

We might find out more early next week.

 

[spsalso]

Looks to have happened at one end of the siding at Buelow, Montana (about a mile west of Joplin, Montana). I can't yet figure out which end.

I expect the track switch underneath the train is involved.

A bit later: it's looking like it was the west end of the siding, and that the train went through the switch in a trailing manner (from the two-track end).



spsalso

 

[timeforchili]

based on the photos from that foxnews link this looks likely to be track caused due to a sun kink, or track buckle. The first few cars will usually make it over the buckle, often deforming it more as they pass, and the cars behind derail and get dragged along for as long as they decide to hold on.

some other photos make it look like it derailed well before the switch.

https://pbs.twimg.com/media/FAQflwMVgAIVTJN?format=jpg&name=large

ignore the annotations on there, but best photo of the whole scene. all that track damage/warping likely occurred after the initial derailment.

it was much hotter on 9/25 than prior days, peak temperature right when they derailed.

https://www.wunderground.com/history/daily/us/mt/great-falls/KGTF/date/2021-9-25

there are ties staged for installation by a tie gang alongside the track. These ties are likely staged for several miles in either direction. This tells investigators a few things, for certain that the track is in poor enough condition to warrant a major rehabilitation. It could also point to track disturbance as a factor if the tie gangs had started work in the area ahead of where the cars stopped. If they did, this will become the primary focus of the NTSB. the track disturbance caused by installing new ties and resurfacing the track reduces its ability to resist the thermal forces and make it more likely to buckle. local maintenance gangs will also often do spot work in areas like this to try and limp it along until the tie gang arrives for the full rehab. in hot weather both categories of that work require slowing the track down after completion until some tonnage is hit, but not always in normal weather.

BNSF/NTSB likely knew the primary cause within their first 20 minutes on site but will deep dive all records on inspections, hot weather inspections, defects, rail cuts, repairs, etc.

 

[MacGyverS2000]

Is 84F really that bad? Seems like they would have plenty of much hotter days to expand the tracks, so an 84F day would be relatively mild in comparison...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[spsalso]

it was much hotter on 9/25 than prior days, peak temperature right when they derailed.

And on 8/1, it was much hotter than on 9/25. Seems there should have been a sun kink then, too. A bigger one, I think.


spsalso

 

[1503-44]

84°F is relative. Installation temperature might not have been 20°C 68F

 

[spsalso]

84°F is relative. Installation temperature might not have been 20°C 68F

Railroad track people are familiar with thermal expansion and how to plan for it during installation at various temperatures. For over a century, I think.


You might be suggesting that this section of track was installed after 8/1/2021 (see my comment above), and done improperly.

Certainly a possibility.



spsalso

 

[caenright]

For that part of the country, Rail Neutral Temperature/Rail Laying Temperature is likely about 95F, but the threshold for a speed restriction was probably 85F. While close to that temperature threshold, it's not out of the question for a kink to develop from a locally warmer area. If temperatures were above whatever threshold exists for BNSF in that area, there should have been an inspection performed to determine if rail was tight, and appropriate restrictions placed.

I tend to agree with the sun kink explanation, and this is logical given the unseasonably warm conditions overall and that the CWR was likely laid some time back with a lower RNT. Since this was close to a turnout for a siding, I can see how there would be more lateral restraint on the rails and potentially a higher likelyhood of kinking with temperature increase.

The UPRR "Field Handbook" has a detailed section on CWR management, including some good pictures of sun kinks that are exacerbated at structures and turnouts. See

https://www.up.com/emp/engineering/...ook/Complete Book/Complete Field Handbook.pdf

-----------------------------
Chris Enright
PE, Colorado
Roads and Trains

 

[spsalso]

And the reason there was a heat kink at 84 degree when there wasn't one at 96 degrees 2 months earlier is.....

Thank you very much for the Handbook.

I am a recent beneficiary of another download, which I will share:

NTSB Accident Report--CSX Autotrain April 18, 2002

It's an NTSB accident report. It makes interesting reading, especially when compared to the UP book.

spsalso

 

[JohnRBaker]

I saw in that UP handbook where they warn that the actual temperature of the rails, on a sunny day, can actually be as much as 30 degrees higher than ambient.

John R. Baker, P.E. (ret)
EX-'Product Evangelist'
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[timeforchili]

spsalso - i think this will be entirely dependent on what track work occurred in the days leading up to this event. the neutral rail temp in montana is probably 95, which would mean heat restrictions should start at 85 deg ambient temperature so definitely in range of problematic temperatures. if the tie gang installed ties on the cooler days before 9/25, they would've done so without the heat restriction rules of replacing no more than 50% at a time. if they did 50% replacement the day before, then that track structure would be extremely compromised on the hot day.

on the 95 degree days in august they likely had blanket speed restrictions out, along with daily heat inspections looking for tight rail conditions. they also would have been pretty careful about any maintenance work that would disturb the roadbed at that temperature.

in a former life i did track maintenance in the central valley of CA, 100+ degrees all summer. we'd never see a tie gang later than march. broken rails all winter long because we were more afraid of sun kinks and would always remove extra rail.

 

[timeforchili]

NTSB Vice Chairman says train derailed before reaching the switch, as suspected.

https://billingsgazette.com/news/lo...cle_11cd6bfe-a5dc-51eb-963a-6510af81f72f.html

"The derailment occurred on a gradual right hand curve, "and it occurred prior to reaching the switch that is there for a siding," Landsberg said.
looks to be here:

https://goo.gl/maps/o9jzRgGvgEcETsd27

also, bnsf says the ties were for the siding, not the main:
Toward the end of the press conference Landsberg was asked about railroad ties along the track before the derailment.

"That will be one of the questions that we look at, I believe. Maintenance is going to be a very big concern for us," he said. "We don't know at this point exactly what happened, whether it was a track issue, whether it was a mechanical issue with the train. So, all of these things are open."

Kent, with BNSF, later said that the ties referenced were for a project on adjacent siding and not on main line where the derailment occurred.

 

[FacEngrPE]

NTSB Investigators on Scene at Montana Amtrak Derailment

No public docket yet.

 

[spsalso]

As of November 6, 2021:

Link

This is the report after aproximately one month of NTSB being on the scene:

Link

All information in the report was available within a few days (closer to hours, really) of the crash. This is the first time I can recall where the investigative crew found nothing to report in a month's work.

The report did not comment on that lack.


spsalso

 

[JohnRBaker]

That first link goes to a site that states that the document is not yet available.

John R. Baker, P.E. (ret)
EX-'Product Evangelist'
Irvine, CA
Siemens PLM:
UG/NX Museum:

The secret of life is not finding someone to live with
It's finding someone you can't live without

 

[waross]

A simple question yet to be answered.
Was it a track failure or a train failure?
A broken wheel or suspension failure of some sort.
Have I missed something?

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[spsalso]

"That first link goes to a site that states that the document is not yet available."

That was my point.

"Have I missed something?"

I don't think so. It appears NTSB worked on this for a month and haven't come up with anything to report.

Neither was there a press conference concerning this report. Thus no questions could be asked.


spsalso

 

[waross]

I feels like a diagnosis I heard often in Central America;
"It broke itself!"

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[phamENG]

Due to the possible complexity of railroad accidents and the extensive nature of the investigation process, a railroad accident investigation often requires 12 to 24 months to complete. Depending on the circumstances of the accident, the IIC along with other NTSB investigators and party members may travel to the accident site to collect information. The NTSB has the authority to designate organizations, corporations, or agencies as parties to the investigation. Only representatives who can provide technical or specialized expertise are permitted to serve as a party member and report directly to the NTSB. The NTSB will not determine the cause of an accident while on scene.

The following are reports and products that can be issued during the investigation:

The preliminary report provides a synopsis of factual information collected during the on-scene phase of the investigation. This report is typically available a few weeks after the accident.​
The final report provides a summary of relevant factual information, analysis, findings, probable cause, and safety recommendations issued or reiterated as a result of the investigation.
​An accident brief consists of a short accident narrative, probable cause, and if issued, safety recommendations. Briefs are normally limited to factual information but may contain analysis.

​

The report or brief replaces the preliminary report and is typically available 12 to 24 months following the accident.

​You can access NTSB reports by searching the NTSB’s Case Analysis and Reporting Online Search (CAROL).

An accident docket contains relevant factual information collected during an accident investigation that is used in the development of reports, briefs, and safety recommendations. The docket is available to access at an appropriate time during the investigation, or at its conclusion.

Many investigations result in safety recommendations or safety alerts. The NTSB can issue these at any time during the investigation process.​​

It happened what? 2 months ago? The NTSB doesn't move that fast, and they don't claim to. The preliminary report is up, and according to the standard time line the Final Report won't be available until September of 2022 or possibly even 2023. While I hope it doesn't take that long, expecting it to be done in a month and a half is a bit...optimistic? Yeah, we'll go with that. Optimistic.