Ohio Train Fire 4/2/23

[ainsworth74]

Precision Schedule Railroading (which is neither precise, scheduled nor indeed really railroading) strikes again. The state of Class I railroads in the US really is quite abysmal and yet I fear that even this disaster won't be enough to shake anyone into the urgent need for reform of the sector (which if they won't do it themselves will need to be done to them).

 

[eldomtom2]

Even allowing for the much greater numbers and lengths of freight trains in the US and Canada in comparison with Europe, their tolerance of frequent major derailments does surprise me. At a guess I'd estimate that a major derailment (10 or 20 wagons not just off the road, but piled in heaps, down in rivers, on fire, etc.) occurs pretty much every week on average. RJCorman* seem to have a large fleet of side-boom tractors on standby across most of the country, ready to go in to clear up the mess. They're very quick to rebuild and reopen tracks, no doubt in part because they get so much practice.

(*other contractors are probably available)

America tolerates a much lower degree of rail safety than the UK - for an example, they have had only one year without double-digit staff fatalties, and that's with a workforce that isn't even double the size of the UK's...

 

[ac6000cw]

Workers for Norfolk Southern are now admitting that they had concerns about the train’s size before the accident and that those concerns probably contributed to both the derailment and an earlier failure after the train left Illinois on February 1, according to CBS News. One employee tells the news source that “we shouldn’t be running trains that are 150 car lengths long,” and that “it’s highly probable the impacts of the derailment would have been avoided” if the train hadn’t been so heavy. The train might pose a safety danger, according to two employees who talked with Motherboard.

The former director of the Federal Railroad Administration, Sarah Feinberg, agrees that the size of the train would have been a concern, Wayne Dupree noted.

“I was not satisfied with the lengths of the trains, and they were 80 or 90 cars long,” she says of her tenure at the FRA. The derailed one had 151 cars. Even while adding more cars might make trains more efficient overall for railroad firms, it takes more time for staff to examine such a train. Shorter inspection times in the name of such efficiency, according to Jared Cassity, national legislative director for one of the Norfolk Southern workers unions, made it likely that the car that derailed earlier this month hadn’t been examined “in some time.”

The accident, which was followed by a controlled burn that released a toxic plume into the sky, was, in his words, “expected” given the extra length and weight as well as the presence of all of the hazardous materials. “That will happen again if nothing changes.”

The railroad company has stated that the length of the train was known, managed, and should have been stable. According to the company, the weight distribution of this train was uniform all throughout, and it had a locomotive in the middle of the train that “helps manage the dynamic forces of the train and reduces occurrences of broken knuckles,” according to a statement from Norfolk Southern to WKBN. Norfolk Southern also notes that some of the claims being made about the train’s size are “simply false.”

Not really sure what's notable about a train "consisting of more than 150 cars, 9,300 feet in length, and 18,000 tons" in North America - 70 years ago Norfolk & Western (one of the major predecessors of Norfolk Southern) was running 200 car loaded coal trains with steam power, and 8,000 to 10,000 feet passing loops have been commonplace for many years.

Although AFAIK the train involved in the accident was a mixed freight from a yard in East St. Louis to Conway Yard near Pittsburgh, most bulk commodity trains (e.g. coal, grain, oil, ethanol, other minerals) will be 15,000 to 20,000+ tons and 120+ cars long and have been for many years. Mixed freights can tend to be longer than bulk trains because the average weight per car is lower (some will be empty or loaded with low-density cargo) - I think the practical upper length limit is more about the difficulties/time taken in assembling and breaking up very long trains in yards, passing loop length limitations and excessive brake propagation delays/leakage (especially in very cold weather).

Hopefully in the wake of this accident at least the train consist rules will be revised to limit the numbers of high-risk hazmat cars in a train, and also that they be more distributed within it i.e. more non-hazmat cars between them.

(If you want 'pushing the limits' railroading in terms of train operations, try Canada... 20,000+ tonne bulk trains over Kicking Horse Pass and Rogers Pass out west, 2.2% gradients, -25C or lower temperatures, snow, ice, avalanche risks...and that's the original land of 'Precision Scheduled Railroading'. Also 30,000+ tonne, single-manned iron ore trains over on the east side.)

 

[eldomtom2]

passing loop length limitations

My general impression has been that freight railroads in the US do not care about running trains too long to be looped.

 

[ac6000cw]

My general impression has been that freight railroads in the US do not care about running trains too long to be looped.

It's OK if one of the trains (being crossed or overtaken) is short enough to fit in the loop. Fun and games start when they are both too long e.g. having to do a 'Double Saw-by'...given the disruption that can cause it's not something they'd choose to do. I suspect a railroad dispatcher that allowed it to happen (if it was within their control) might have an uncomfortable meeting with their supervisor.

(In the UK, back in manual signaling days, the Felixstowe branch crossing loops became too short to cross two long container trains - they could only cross with passenger/short freight trains. When BR re-signaled Ipswich East Suffolk Jn. - Felixstowe, the loop at Derby Road was considerably extended to fix that problem)

 

[eldomtom2]

The EPA have taken over and announced their intent to get NS to pay for the cleanup in full: https://www.npr.org/2023/02/21/1158532207/epa-east-palestine-train-derailment

The NTSB have issued a prelimary report: https://www.ntsb.gov/investigations/Documents/RRD23MR005 East Palestine OH Prelim.pdf

 

[ac6000cw]

The NTSB have issued a prelimary report: https://www.ntsb.gov/investigations/Documents/RRD23MR005 East Palestine OH Prelim.pdf

So a hot axle box/bearing failure, where the temperature was only high enough to trigger a critical warning over the radio at the hot-box detector just before the derailment site (the temperature at the previous detector, 19 miles back, was well below the required threshold). The crew were already braking to a stop (using dynamic braking) because of the HBD warning when the derailment occurred and put the train into emergency braking.

(US hot-box detectors normally 'talk' to trains over the standard railroad voice radio system, so if everything is OK they'll say something like "Detector milepost 123, axles 512, no defects, 3:45 PM" just after the train has passed. I once experienced a train I was 'chasing' get stopped by one, but can't remember what the detector said - the train continued at low speed after a while).

 

[John-H]

Long trains have a tendency to bunch up and then pile up when the emergency brake is applied. If the train slack is already compressed due to the use of the dynamic brake, it could be worse.

John

 

[furnessvale]

Perhaps more sophisticated power distribution is needed, rather than 5 or 6 locos at the head end as often seems to occur?

 

[ac6000cw]

Long trains have a tendency to bunch up and then pile up when the emergency brake is applied. If the train slack is already compressed due to the use of the dynamic brake, it could be worse.

John

It's inevitable really, due to the propagation delay of any change in brake pipe pressure. Emergency applications propagate faster (by design) because when brake control valves on a vehicle detect a brake pipe pressure drop to well below the normal full-service application pressure they locally vent the brake pipe. I was once watching a freight go by when it went into 'emergency' and remember hearing the loud 'psst' noise propagating along the train.

Perhaps more sophisticated power distribution is needed, rather than 5 or 6 locos at the head end as often seems to occur?

With modern high-power/high tractive effort locos, more than 4 on the head-end (under power) is unusual these days - the in-train forces get too large if the train is heavy enough (especially the AC-drive locos - a pair of those can generate tractive effort roughly the same as the coupler strength). I suppose if you had a real hot-shot intermodal that needed loads of power at speed, more than 4 on the head-end might be OK (with care) but even then putting the extra locos on the tail makes just as much sense. Much the same 'rules' apply in relation to dynamic braking i.e. no more than n-axles of it at the head-end.

I think Locotrol (the standard DPU radio control system) allows for up to 4 or 6 'slaves' per train, but of course distributing say 2 x head end plus 4 x DPUs as individual locos along the train would take a lot longer to set up than having 2 head plus 2 x DPU pairs or 3 head plus 1 x DPU triplet (1+1+1+1+1+1 vs 2+2+2 vs 3+3). The most 4400hp locos I've personally seen (working) on a train is seven in a 3+2+2 formation, on a very long eastbound BNSF intermodal en-route to Cajon Pass. On the same trip, I twice saw a very long & heavy UP mixed freight climbing Tehachapi Pass (slowly) with six matched AC-drive ES44AC locos in a 3 head + 3 mid formation - pretty impressive when you think about the amount of tractive effort being generated...

 

[43096]

It's inevitable really, due to the propagation delay of any change in brake pipe pressure. Emergency applications propagate faster (by design) because when brake control valves on a vehicle detect a brake pipe pressure drop to well below the normal full-service application pressure they locally vent the brake pipe. I was once watching a freight go by when it went into 'emergency' and remember hearing the loud 'psst' noise propagating along the train.


With modern high-power/high tractive effort locos, more than 4 on the head-end (under power) is unusual these days - the in-train forces get too large if the train is heavy enough (especially the AC-drive locos - a pair of those can generate tractive effort roughly the same as the coupler strength). I suppose if you had a real hot-shot intermodal that needed loads of power at speed, more than 4 on the head-end might be OK (with care) but even then putting the extra locos on the tail makes just as much sense. Much the same 'rules' apply in relation to dynamic braking i.e. no more than n-axles of it at the head-end.

I think Locotrol (the standard DPU radio control system) allows for up to 4 or 6 'slaves' per train, but of course distributing say 2 x head end plus 4 x DPUs as individual locos along the train would take a lot longer to set up than having 2 head plus 2 x DPU pairs or 3 head plus 1 x DPU triplet (1+1+1+1+1+1 vs 2+2+2 vs 3+3). The most 4400hp locos I've personally seen (working) on a train is seven in a 3+2+2 formation, on a very long eastbound BNSF intermodal en-route to Cajon Pass. On the same trip, I twice saw a very long & heavy UP mixed freight climbing Tehachapi Pass (slowly) with six matched AC-drive ES44AC locos in a 3 head + 3 mid formation - pretty impressive when you think about the amount of tractive effort being generated...

This video from the late 1990s has nine 4400hp locos powering (all either AC4400CW or AC60/44 convertibles with 7FDL engines): three at the front, four more mid-train and two more pushing at the rear.

 

[Gag Halfrunt]

It's inevitable really, due to the propagation delay of any change in brake pipe pressure.

Which brings us to electronically-controlled pneumatic brakes.

Rail Companies Blocked Safety Rules Before Ohio Derailment

Norfolk Southern helped convince government officials to repeal brake rules — and corporate lobbyists watered down hazmat safety regs.

www.levernews.com

The sequence of events began a decade ago in the wake of a major uptick in derailments of trains carrying crude oil and hazardous chemicals, including a New Jersey train crash that leaked the same toxic chemical as in Ohio.

In response, the Obama administration in 2014 proposed improving safety regulations for trains carrying petroleum and other hazardous materials. However, after industry pressure, the final measure ended up narrowly focused on the transport of crude oil and exempting trains carrying many other combustible materials, including the chemical involved in this weekend’s disaster.

Then came 2017: After rail industry donors delivered more than $6 million to GOP campaigns, the Trump administration — backed by rail lobbyists and Senate Republicans — rescinded part of that rule aimed at making better braking systems widespread on the nation’s rails.

Specifically, regulators killed provisions requiring rail cars carrying hazardous flammable materials to be equipped with electronic braking systems to stop trains more quickly than conventional air brakes. Norfolk Southern had previously touted the new technology — known as Electronically Controlled Pneumatic (ECP) brakes — for its “potential to reduce train stopping distances by as much as 60 percent over conventional air brake systems.”

But the company’s lobby group nonetheless pressed for the rule’s repeal, telling regulators that it would “impose tremendous costs without providing offsetting safety benefits.”

That argument won out with Trump officials — and the Biden administration has not moved to reinstate the brake rule or expand the kinds of trains subjected to tougher safety regulations.

 

[ac6000cw]

This video from the late 1990s has nine 4400hp locos powering (all either AC4400CW or AC60/44 convertibles with 7FDL engines): three at the front, four more mid-train and two more pushing at the rear.

I watched the video without being able to read the title, saw the mixture of SP and UP liveried locos, a coal train, the snow and thought - that must be on the (now closed) Tennessee Pass . With 3.5% gradients and a 10,000+ foot summit they had a good excuse for needing nine locos...

Before Canadian Pacific decided they really needed to spend a fortune on adding a low-gradient line over Rogers Pass in the 1980s, it took twelve (thirsty) SD40-2 locos to get westbound 14,000+ ton coal trains over it. They left Golden with four locos on the head and two mid-train (radio controlled). At Rogers Camp a manned six loco helper set was cut in about 3/4 back along the train - making a 4+2+6 loco formation. 36,000hp for 14,000 tons up a 2.2% gradient. When I saw some of them in 2017, they had 2+1+1 ES44AC/AC4400CW hauling heavier trains with the assistance of 10 miles of tunnel - the power of expensive civil engineering

 

[HSTEd]

Which brings us to electronically-controlled pneumatic brakes.

Rail Companies Blocked Safety Rules Before Ohio Derailment

Norfolk Southern helped convince government officials to repeal brake rules — and corporate lobbyists watered down hazmat safety regs.

www.levernews.com

Ofcourse once ECP braking is adopted the data bus allows you to inexpensively fully instrument every vehicle.
At which point you can dispense with intermittent trackside hotbox detectors and just have temperature sensors on every axle box.

Probably would help here by providing more warning time as well as improving braking performance and control

 

[eldomtom2]

The chairman of the Brotherhood of Maintenance of Way Employes has written an open letter to the Ohioan Governor, alleging the workers employed in cleaning up the crash site were not provided with PPE despite reporting nausea and migraines.

The accident seems to be drawing a spotlight on the safety culture of American railroads, with the Guardian now reporting that freight car inspectors at Union Pacific were given no formal training and were under heavy pressure not to mark cars for repair.

 

[jon0844]

Another derailment in Ohio, this time in Springfield.

Link: I read this on a Tweet, but cannot see the CNN article as we're getting the European site and the link doesn't appear to work.

No hazardous materials or injuries reported.

 

[DelW]

Another derailment in Ohio, this time in Springfield.

Link: I read this on a Tweet, but cannot see the CNN article as we're getting the European site and the link doesn't appear to work.

No hazardous materials or injuries reported.

There is a report by WBNS 10TV on YouTube

Looks like mostly autoracks and boxcars off the track, but it won't help NS's reputation in the state. Residents advised to stay at home, and some have lost electricity due to damage to overhead lines.

 

[eldomtom2]

Probably only news because attention is on the railroads, sadly.

 

[ac6000cw]

Probably only news because attention is on the railroads, sadly.

Yes, agreed.

Ofcourse once ECP braking is adopted the data bus allows you to inexpensively fully instrument every vehicle.
At which point you can dispense with intermittent trackside hotbox detectors and just have temperature sensors on every axle box.

Of course, but my guess is that it's mixed freight trains (like the one involved the East Palestine derailment) will be the last ones to benefit, due to the variety of freight cars and car owners involved. Block/unit trains tend to operate in a merry-go-round fashion so it's much easier to keep ECP-equipped cars and locos together.

Probably would help here by providing more warning time as well as improving braking performance and control

ECP should propagate the emergency application (after a train has parted) much faster, but a 150 car train running at speed has a huge amount of kinetic energy to dissipate before everything stops - so it's pretty much inevitable that the rear part of a train will push a fair number of cars off the track once one has derailed. At East Palestine the car with the failed bearing was 23 cars back, so there were 126 cars + a loco (maybe 10000+ tons?) behind it running at 47mph... 38 cars derailed in total, according to the preliminary NTSB report.

 

[eldomtom2]

The NTSB is opening an investigation into the safety culture at Norfolk Southern, after both this derailment and a series of worker fatalities.

Unfortunately I think part of the reason for the poor safety culture of American railroads is the NTSB itself - half the time they write employee deaths off with a "should have followed the rules".

Nor, of course, do I think any problems they will find with Norfolk Southern will be unique to it.

 

[ac6000cw]

The NTSB is opening an investigation into the safety culture at Norfolk Southern, after both this derailment and a series of worker fatalities.

As the NYT article requires a login, this is the original NTSB press release - https://www.ntsb.gov/news/press-releases/Pages/NR20230323.aspx

 

[ainsworth74]

Nor, of course, do I think any problems they will find with Norfolk Southern will be unique to it.

Indeed, it seems fairly clear that this is likely systemic amongst the Class 1s it's just NS having a time in the spotlight and/or a bad patch. But will the NTSB actually go looking at the other railroads? The press release certainly doesn't suggest they will!

 

[eldomtom2]

The Department of Justiice is now suing Norfolk Southern under the Clean Water Act.

 

[eldomtom2]

The FRA have completed their investigation into Norfolk Southern's safety culture. It's very gentle considering its fairly serious findings, some choice quotes:

Further, reviewing the maintenance and inspection history of the territory identified a failure to prioritize critical safety work necessary to ensure safe operation. FRA observed that NS did not have consistent management oversight of the territory nor staff to perform critical repairs, noting many instances when NS maintainers were covering multiple territories and NS had multiple vacancies in its signal management overseeing work in the territory.

However, an almost equal percentage of employees reported that safety continued to be a lower priority than production and train movement. A handful of employees (approximately 5) indicated that safety was listed as the fourth priority in documents that listed NS’ responsibilities.

Of note are comments provided from labor leaders in semi-structured interviews as well as in about 10% of comments from survey interviews that many employees across the NS system, crafts, and years of experience do not report concerns to management out of fear of retaliation or disciplinary action.

However, it should be noted that about one-third of labor leaders interviewed, as well as one-third of craft employees responding to the survey, expressed a reluctance to stop an unsafe action. Reasons for this included fear of retaliation, fear of disciplinary action, fear of punitive action taken against an employee if there was a disagreement between labor and management if an action or situation was unsafe, and perception that production must come before safety.

Overall, a general theme emerged from the focused inspections performed as part of this safety culture assessment: a lack of trust between employees and management. During FRA’s signal focused inspection, FRA observed poor teamwork and leadership in the territories inspected. Specific findings to support this included a territory without consistent frontline supervision for 18 months, and observations of employees performing maintenance tasks for which they had not received proper training. Further, there is discontent with NS’ training program and a general feeling by craft employees that NS is not committed to ensuring employees are trained and prepared to safely perform their tasks. This perception by employees that NS is not providing them with the resources needed to perform their jobs safely is an example of this lack of trust.

 

[mr_moo]

Wow. Damning. Sad (but unfortunately not surprising) that such cultures still exist. One only has to look at some of our own UK industry failings to see evidence of that but thankfully it's a much less common thing than it used to be (I think/hope).

 

[eldomtom2]

Wow. Damning. Sad (but unfortunately not surprising) that such cultures still exist. One only has to look at some of our own UK industry failings to see evidence of that but thankfully it's a much less common thing than it used to be (I think/hope).

I can say that the way railway workers here talk about safety is very different to the way I've seen American rail workers online talk about safety - there the attitude in some places seems to be still "rules are to protect the company, not you; only follow them when management is watching lest you get yelled out for slowing things down".