Hacker News Reader

Comment by thechao | Hacker News Reader

Has anyone seen a good analysis on what went wrong? I found a USAToday article with a partial fragment saying that one of the wheels had a failed bearing, and this was the proximal cause of the derailment. It seems to me that (as pointed out in the article) the mean-time-to-failure increases with the length of the train; perhaps nonlinearly? I think the failure rate of bearings is proportional to the length of the train, but you can't "amortize" failure: any failed bearing would lead to a catastrophic error?

To my mind this means that things like ECP (which have been removed) might've fixed the issue, but the fundamental problem is the train length, which has been discussed (in all of its negatives) here, before.

In that sense, the train company is directly at fault: they are engaging in unsafe practices in order to remain profitable, and the practices are fundamentally not safe-able (to make up a word).

View on HN

There are thousands of derailments every year.

The bigger problem is not the derailment but the fact that they released and burned off the vinyl chloride in order to expeditiously clear the tracks and get them running again, at great health and environmental cost to the local community and even several states in the windshed to the northeast (PA, NY, VT)

They burned off the vinyl chloride because the combustion products are relatively benign, it was much safer to remediation personnel to ignite the chemicals in a controlled burn vs having them work around flammable chemicals for weeks, vinyl chloride is quite volatile and acute exposure to high concentrations of vinyl chloride vapor is dangerous, and they wanted to get everyone back in their homes (and yes, get the railroad back up and running) as quickly as safely permissible.

>the combustion products are relatively benign

Who told you that? Phosgene and dioxins are "relatively benign"? Dioxins might more properly be called forever chemicals, and will contaminate the soil and water of five states for decades.

I understand they were trying to avoid a bleve. The solution they chose was not based principally on minimizing citizen harm though

I haven't seen anything concrete about dioxins coming from the burning of vinyl chloride. From what I've seen, the known dioxin problem is from the burning PVC, which was in bulk cars that were ignited by the derailment (as opposed to deliberately set on fire).

FWIW Only one vinyl chloride car was listed as leaking. I would think the alternative was to bring in mobile pumps, new tanker cars, and possibly chillers. And that should probably be made the default standard, rather than defaulting to "controlled" release in an emergent situation.

Someone else a few days ago linked a paper that measured minuscule amounts of dioxins from burning vinyl chloride. Minuscule as in measurable, but probably not a concern.

Phosgene also has very low environmental persistence. I think the actual amount of dioxin created from combustion of vinyl chloride monomer at atmospheric pressure is very low. In any event, both were dispersed into the atmosphere at concentrations that pose(d) very little danger to people outside of the temporary evacuation zone.

I stand corrected. They did a great job and no one in East Palestine is sick

That is correct. Not a single person in East Palestine was hospitalized.

I just wanted to follow up here. Yes, people have been hospitalized. Would you care to revise your statement?

https://twitter.com/StatusCoup/status/1629530227146731521

Correct. There is no one in East Palestine who got sick from this.

There's many derailments throughout the country and chemical spills of other sorts have happened many times.

> There is no one in East Palestine who got sick from this.

You're saying this a decade or two early, don't you think? Cancer rates are a trailing indicator.

I'm not going to try to address the normalization of deviance implicit in the rest of your comment, because what would be the point?

None of the contaminants are carcinogenic, really

> Vinyl chloride is a mutagen having clastogenic effects which affect lymphocyte chromosomal structure.[20][22] Vinyl chloride is a Group 1 human carcinogen posing elevated risks of rare angiosarcoma, brain and lung tumors, and malignant haematopoeitic lymphatic tumors.[23] Chronic exposure leads to common forms of respiratory failure (emphysema, pulmonary fibrosis) and focused hepatotoxicity (hepatomegaly, hepatic fibrosis). Continuous exposure can cause CNS depression including euphoria and disorientation. Decreased male libido, miscarriage, and birth defects are known major reproductive defects associated with vinyl chloride.

Yes, they found elevated rates of some cancers in people who spent their careers in poorly ventilated vinyl chloride production facilities. For the sort of concentrations and durations we’re talking about in East Palestine, it’s not carcinogenic.

It's nice to see someone using some logic in here, there is a ton of ignorance. The only issue is that people always need to find someone to blame. Most aren't going to take facts like you have provided and use that in their assessment of the situation. My main concern would be the groundwater if anything, not the byproducts of combustion that have long blown away in the wind, or the fact Joe Biden didn't come walk around the town. I can imagine the town hall meetings are full of ignorant fools who just want to blame and be angry or political.

My first job (1983) was as a bench tech, at a defense contractor (microwave spy stuff/TEMPEST).

We all had these little bottles on our benches, with a funny silver lid that acted a bit like a "birdbath"[0]. You pumped it a couple of times, and the contents would come up, into a shallow bowl. You'd then take a long "Q-Tip" swab, dip it in the liquid, and use that to clean flux off the PCB you were fixing.

They also had open buckets, and people would wash entire assemblies off, in these buckets.

The liquid was trichlor[1]. We had gallons of it, sloshing around. If you got any on your hand, it made the skin turn white, and flake off. It smelled like the demon spawn of acetone and gasoline.

We were assured that it was completely benign, and not to worry our pretty little heads over it.

Let's just say that it can take some time to overcome [in]vested bias.

[0] https://www.aliexpress.us/item/2251832673900981.html

[1] https://www.cancer.gov/about-cancer/causes-prevention/risk/s...

1-1-1 Trichloroethane???

Even in 1990 or so when I first came across it (also for cleaning PC boards), it was known to be hazardous.

Yup. This was in 1983.

Lots of pretty hairy stuff, in that lab. The gold-plating room had 50-gallon drums of cyanide.

People's dogs and cats just up and died. Last time I checked humans are also mammals that breath.

Yes the pets were present while the controlled burn was taking place. People were evacuated.

The chemical was especially toxic to certain types of animals (notably fish) but much less toxic to humans.

Here's a video on that. https://www.youtube.com/watch?v=b19lpkjUBrM

All of that smoke and no one got sick ? Hard to believe

They evacuated people first.

Wasn't the mandatory evacuation order only for a 1-mile radius?

How much larger are you suggesting it needed to be, and based on what?

Since hot smoke travels upwards, I don't know what making the zone wider would have accomplished.

I think the radius of the evacuation zone should be up to the issuer to justify.

If you want a 1-mile zone it seems that you should be able to articulate why or else it should just get denied. However, it seems that wikipedia et al don't seem to have a citation as to how 1 mile was picked so it's hard to say if a smaller or larger zone should've been done.

Edit: Found a chart from Ohio.gov [1] showing impact areas. It does seem to be much smaller than 1mi.

> Since hot smoke travels upwards, I don't know what making the zone wider would have accomplished.

I mean if hot air only went up the Earth would have no atmosphere after a sunny day.

[1]: https://governor.ohio.gov/media/news-and-media/east-palestin...

I don't want to downplay anything that happened here, but if you said a train of dihydrogen monoxide derailed and the vapors were spreading through the community, some number of people would come forward claiming to be sick. Statistically, someone is going to wake up with a cold on the day of the derailment. And the brain is a very powerful drug, if you scare someone with a chemical spill, they can show symptoms regardless of whether or not some chemical is acting on their body.

This is a risky comment to make, because it sounds like I am supporting NS's cleanup plan, and I'm not. I'm just saying that people saying they're sick doesn't mean the plan was executed incorrectly.

>no one in East Palestine is sick

Gastrointestinal bleeding

https://twitter.com/StatusCoup/status/1626288682326958083

Family including three year old child diagnosed respiratory infections

https://twitter.com/NewsNation/status/1625679079520325633

It appears the family caught a cold

My friend from East Palestine took a video of the water in a creek that runs outside his house. The water is rainbow like an oil spill. This was yesterday. Whatever they combusted and dispersed, it's very much still there and I sure wouldn't want to be drinking that water.

Probably a biofilm.

If it's the same photo, a biologist weighed in on it on Twitter and suggested exactly that.

I mean, i imagine that whatever contamination that is, is separate from the combustion products that went into the atmosphere.

that sounds like iron oxidizing bacteria, did it look like this?: https://stock.adobe.com/search?k=%22iron%20bacteria%22

Maybe this is out of date, but:

> There is also the theoretical possibility of burning vinyl chloride forming dioxins which are known carcinogens. So far, no dioxins have been detected.

https://www.mcgill.ca/oss/article/health-and-nutrition-histo...

Have they tested for dioxins? Can you share?

Yes, they have been testing and sharing with plenty of news outlets if you care enough to seek it out. The chemicals were burned off and have now dispersed, there is no more source to worry about for anything airborne. It's too bad this is an area of mostly poor, uneducated, right wingers that don't probably know much about chemistry or analytical thinking. Perhaps they could put their energy into ensuring the groundwater does not become contaminated instead of carrying on about things that don't matter, such as Joe Biden and air pollution.

https://youtu.be/L1I6eVP8QQk

Rivers are filled with something that’s heavier than water. Fish are all dying.

Same from a video a friend took yesterday, but he didn't even need to disturb the creek, it just had the sheen on top right there.

Fish are extremely sensitive and die off all the time from levels of contamination that don’t pose much concern for people.

iron oxidizing bacteria often produce that sheen when disturbed

one way to tell is if the sheen, when itself disturbed, stays broken into pieces (bacteria) or rejoins together (oily)

>it was much safer to remediation personnel to ignite the chemicals in a controlled burn vs having them work around flammable chemicals for weeks

It would have taken weeks for a large org like Norfolk Southern to unload the payload to other containers? Seems like an unreasonably long amount of time.

> the combustion products are relatively benign

If you consider phosgene and HCl to be "benign" based on their half life sure, but this assumes complete combustion. The cloud of visible black smoke emitted shows that clearly didn't happen. Partially combusted vinyl chloride monomer is some horrible mix of carcinogens that will contaminate the area for years.

Nah

>and even several states in the windshed to the northeast (PA, NY, VT)

My understanding is that contemporaneous weather radar[0] put the plume southeast at the time of the burn.

[0] https://www.wdtn.com/news/ohio/east-palestine-train-derailme...

That is good info, thanks for sharing.

Here is the model that NOAA created - and which has not been officially released to the public yet: https://twitter.com/FalconryFinance/status/16258723275684208...

Thanks, I had not seen that.

On the journalism side, the source seems to be extremely alarmist,[0] so I doubt much skepticism was practiced there.

On the NOAA side, it would be interesting to see the full set of assumptions behind this model. I would presume the vast majority of emissions would occur during the burn itself (February 6th, afternoon), but this model seems to assume a significant (undiminished?) outflow at midnight on February 8th.

"Garbage in, garbage out."

[0] https://twitter.com/FalconryFinance/status/16258752974837637...

> There are thousands of derailments every year. The bigger problem is not the derailment [...]

I think that's a problem worth discussing as well. Obama passed legislation to upgrade trains braking systems to make them a lot safer, which was then weakened by lobbying, which was then repealed by Trump, then the rail companies spent billions in stock buybacks and tried to tighten the screws on their labour force even further for cost-cutting reasons.

There are a lot of pathological behaviours that have gone into "thousands of derailments per year" that are worth discussing, and I'd say those are a lot more important than just this one derailment.

The way to put the screws to the rail companies isn’t to regulate better brakes - just have the NTSB have a policy of shutting down any line that has a derailment for months or longer. The companies will solve it themselves.

The rail companies are "too big to fail" since they are core national logistics.

Then maybe they shouldn’t be private companies that cut corners for profit putting both people and that logistical operation at risk.

>they are core national logistics

If they cannot operate safely and they are essential to our national security, perhaps they should be nationalized. I bet the threat of nationalization would get their collective asses in gear.

They will raise prices and if other logistic will fill the gap, well, they weren't so core after all

Fine them millions per derailment. Increasing with each one.

This can work but they can fight fines, or eat it as cost of business. Much harder to argue against a safety-shutdown.

> Obama passed legislation to upgrade trains braking systems to make them a lot safer, which was then weakened by lobbying, which was then repealed by Trump

I'm not sure it was political. The FAST Act, the legislation from 2015 that mandated that regulators require ECP brakes on some trains also required that the National Academy of Sciences produce a report on the effectiveness of such brakes which would then be used to review the regulation to see if the assumptions behind it were valid. The regulators then had to either publish an explanation of how the regulation was justified or repeal it.

In 2017 the NAS report came out, and found that there really wasn't much evidence either way, and said they were unable to conclude that ECP brakes would outperform other types of brakes.

Given that NAS report I am not at all sure that the outcome would have been different under a Clinton administration.

> fact that they released and burned off the vinyl chloride in order to expeditiously clear the tracks and get them running again

I was under the impression that there was risk of a Boiling liquid expanding vapor explosion which would have been even more catastrophic than the "controlled" release and burn off.

We know the bearing was on fire for 20+ miles before it derailed, and it passed through at least one hotbox detector in that condition without being detected. On average, such detectors are placed every 10-20 miles. The one in East Palestine spotted the problem but it was too late to prevent the derailment.

Maybe a start would be to adjust regulation such that detectors need to be placed with higher frequency. ECP is a nice idea, but that's a last ditch effort and not a generalized solution.

Do we know it was without being detected? I got the impression somewhere that it was ignored because they frequently are for a while due to lack of sidings of adequate length to stop in.

The length of the siding shouldn't prevent you from using it. It should just require the main track to be signalled stopped for other traffic in either direction while you are occupying it during the set out procedure. If the train is long enough, this could be quite a long time if you have to dump and then rebuild working pressure in the brake system.

There’s also the possibility that mechanical parts will seize up (especially in the case of overheated bearings) and will be unable to get moving again if you stop.

That is still better then a derailment of a train carrying toxic chemicals…

I mean, if attempted to apply the same logic to driving your car I would say you should have your license taken away.

They can always split the train and cart off the working cars to a siding, and then if need be drag the sparkler at low speed with little risk of derailment. Or worst, you bring in a crane and replace the bogie on a siding. Or any number of actions that don't start a statewide emergency. Talk about normalizing deviance in name of profits.

Why don't they put the hotbox detectors on the trains themselves? I'm cynically assuming cost?

They won't spring for a system to apply the brakes on all cars at the same time. They sure aren't going to wire every axle up with sensors and network the whole train.

The hotbox detector could have seen the bearing overheating and been ignored, or have malfunctioned or just not been sensitive enough to detect the problem when the train passed. It’s old tech that should have caught this in most cases. The even older tech solution would be adding a caboose on trains carrying hazardous materials. Maybe you don’t need a caboose today but someone watching cameras in a cube could do the same thing.

ECP does not seem like a solution, and addressing how hazardous chemicals are moved is something that needs to be done.

They shipped the wheelset to DC for inspection but we haven't seen the report from that inspection yet. They are also going to inspect the actual site, but are waiting until the site is decontaminated.

So we might get a good ruling on the bearing in the coming weeks but based on other reports that I have read (and I quite enjoy the industrial disaster genre of podcast/video) it will probably be a year or two before the full report is complete.

> are waiting until the site is decontaminated

Right. It's all completely harmless and there's no risk to any residents, but the site is also too contaminated for investigators to make even a very brief visit. Makes sense to me.

This article contains a video of the defective train passing a webcam some miles before the derailment: https://www.post-gazette.com/news/transportation/2023/02/10/...

I got to that page by reading this page: https://e40.github.io/2023-02-18-train-safety.html

https://homesignalblog.wordpress.com/2021/12/27/efficiency-a... for the general issues. Uday is writing a senior thesis on this topic, I believe, which should be a great read.

From what I read the NTSB has not started its investigation because their are waiting for the train cars to finish being cleaned up. After that they will begin and we'll get some preliminary findings and eventually final report

Reply on news.ycombinator.com