Derailment and fire in Quebec

[Gerald Henriksen]

However to me that object looks like a tank car, the 'yellow' is rust and the end further from viewpoint looks to have either collapsed or been blown outwards.  Incidentally what I do find worrying about this incident is the apparent ease with which a fire started (presumably a relatively light oil?) and the fact that a number of cars would appear to have exploded instead of being ruptured.  One can't but hope that this appalling tragedy will cast some light on the design of tank cars and the ways in which their safety could be improved as derailments of such vehicles in North America all too often appears to lead to fires and explosions.

Crude oil, being shipped to a refinery in New Brunswick.

 

The tank car design is apparently known to be an issue in accidents, but industry resistance to the cost of either replacing or upgrading the existing fleet:

http://www.bloomberg.com/news/2013-07-09/tanker-cars-in-fatal-quebec-rail-crash-had-drawn-scrutiny.html

 

As for the fact that cars exploded, once a fire of that size starts the heat will both weaken the steel in any cars inside the fire, as well as increase the pressure inside.  Given long enough, and boom.

 

Also if teh information about the fire at Nantes is correct (i.e. there was a fire on a loco) why on earth wasn't that loco immediately removed from traffic for proper technical investigation - seems a very slapdash of running a railway if that wasn't the case although we are of course again relying on media reports which can be both inaccurate and misleading.

But there was no crew on site to do anything (the train had been parked for the night and the engineer was 11 km away in Lac Magentic.

 

 

[PhilJ W]

Surely the brakes are fail-safe? if air pressure is lost the brakes should come on not off. As for the volatility of the oil it is not the oil itself but the fumes given off that is the problem, this in the past has resulted in the loss of one or two large oil carriers.

[Edwin_m]

The brakes are failsafe in the sense that they will come on if the pipe is vented, and hold the train for at least an hour.  However if there is no source of compressed air the system will leak off and become ineffective, so if a train is left unattended some other sort of brake must be used. 

[dibber25]

Air brakes work by air pressure - the greater the pressure the greater the brake application. They do not fail safe in that respect. If a train is standing for a long period the locomotive handbrakes would be applied. According to Mr. Burkhardt all the locomotives had their handbrakes applied. In that case presumably the combination of train weight and gradient overpowered the handbrakes and the tank car handbrakes should probably have been applied as well.

[Gerald Henriksen]

TSB held a press conference, but didn't seem to have much new to say other than confirming that the train was going faster than the speed limit for that section of track.

 

The Toronto Star has an article on the engineer who parked the train:

http://www.thestar.com/news/canada/2013/07/09/lac_megantic_explosion_engineer_tom_harding_beside_himself_after_disaster.html

[Fat Controller]

Air brakes work by air pressure - the greater the pressure the greater the brake application. They do not fail safe in that respect. If a train is standing for a long period the locomotive handbrakes would be applied. According to Mr. Burkhardt all the locomotives had their handbrakes applied. In that case presumably the combination of train weight and gradient overpowered the handbrakes and the tank car handbrakes should probably have been applied as well.

I don't know anything of North American practice, but in continental Europe, and on Eurotunnel trains, chocks are carried in loco cabs, expressly to be put down in front of the first set of wheels (loco or train) on the 'downhill' end of a train which is meant to be stabled for any length of time. There is a published shot of the first E* to traverse the Tunnel, standing on the Up Fast at Cheriton- careful examination reveals a chock under the front wheel.

[dibber25]

If 500 tons of locomotives with the handbrakes on didn't hold the train, I doubt that chocks would have helped.

[Adrian Wintle]

If 500 tons of locomotives with the handbrakes on didn't hold the train, I doubt that chocks would have helped.

 

No, but a clamp-on portable derail likely would have.

 

Adrian

[brossard]

I'm feeling a bit disturbed to learn that brake application relies on air power being available.  I had assumed that brakes were fail safe in that it required air pressure to release the brakes.  This actually comes close to home for me (literally) since my house backs onto the CN line about 50 feet away.  It's a good thing I don't have any valuable china because every now and then the house shakes whenever trains go by at night (they're supposed to slow down but often don't).  No wonder I don't notice the occasional earthquake we have here.  

 

John

[Fat Controller]

I'm feeling a bit disturbed to learn that brake application relies on air power being available.  I had assumed that brakes were fail safe in that it required air pressure to release the brakes.  This actually comes close to home for me (literally) since my house backs onto the CN line about 50 feet away.  It's a good thing I don't have any valuable china because every now and then the house shakes whenever trains go by at night (they're supposed to slow down but often don't).  No wonder I don't notice the occasional earthquake we have here.  

 

John

They're fail-safe short-term; the problem is when you rely on them to hold a train for more than a couple of hours. Then you have one of those 'I'm sure I left it somewhere around here' moments. What I have seen in the departure fans of some of the bigger French yards is a shore-supply for the brakes to keep the reservoir pressure up- you don't just rely on a few handbrakes (which many mainland European vehicles lack, in the same way as most UK coaching stock) to hold 3,700 tonnes of iron ore in place.

What did surprise me with these tanks is that they appeared to be painted black- crude oil is classed as a Class A product in the UK, due to the amount of volatiles it contains, so oil from places like Holybourne is carried in light-coloured tanks to reduce heat gain from the sun.

Edited July 9, 2013 by Fat Controller

[brossard]

I suppose I'm a bit gob smacked at this because I spent 30 years in aircraft engines.  Designers go through hoops trying to ensure that there are fail safes, redundancy, inspections and maintenance to absolutely minimize any chance of a safety related event.  Obviously still not perfect but the survivability of the SFO event was no miracle.  Anyway, let's hope that the investigation will highlight any design, maintenance and/or operational shortfalls and get them corrected.

 

John

[The Stationmaster]

The brake practice when standing on a gradient does seem to be rather hit & miss to say the least - as do several other things that we do know about such as the apparent fire on the loco at Nantes (which seems to have been a put-it-out, walkaway, and leave it kind of job - without seemingly calling railway attention to the fire or if such attention was called then without the company doing anything about it (sorry, more questions than answers there).

 

Leaving a train on a gradient dependent on an unattended loco keeping the engine running in order to maintain brake pressure sounds to me to be the preserve of idiots  (speaking as someone with more than a little experience of operating on gradients).  Loco engines can stop for all sorts of reasons - fire or no fire - and unless the train is properly secured with handbrakes when the auto brakes leak off it will be unbraked and prone to set off on its own.  I don't know the weights involved but if I were investigating this incident one of my first questions would centre around how many handbrakes had been applied on the stabled locos and train before it was left unattended?  That's a matter of very basic operational safety - whatever country you happen to be in.

 

As for scotches or other fancy devices to stop it rolling away I have never ever come across one that worked that well - scotches can supplement handbrakes when relatively little weight is involved but, say, a single scotch or even several with a heavy train would soon turn into kindling wood.

[Adrian Wintle]

As for scotches or other fancy devices to stop it rolling away I have never ever come across one that worked that well - scotches can supplement handbrakes when relatively little weight is involved but, say, a single scotch or even several with a heavy train would soon turn into kindling wood.

 

A portable derail (if used) would likely have prevented this by derailing the lead loco as it started to move (think temporary catch point). They aren't 100% effective, though.

 

The one thing the TSB timeline did confirm is that the locos ran through the town at the head of the train, so the derailment occurred mid-train.

http://www.cbc.ca/news/canada/montreal/story/2013/07/09/lac-megantic-quebec-train-explosion.html

 

Adrian

[Ozexpatriate]

It does look that way at first glance, but if you look at a wide overview of the scene* there are four or five cars to the east of the main pile-up, as well as the two cars lying on their side on the NE arm of the triangle (wye). There are also train cars on the track to the north. It would appear that the point of derailment was at the wye.

Yes. This Canada TSB photo shows tankers on the south branch of the wye (the intended direction of the train) and in what appears to be a pond south of the wye. No locomotives in sight.

[brossard]

It is horrifying to realise that the bar, where dozens of people were relaxing, was right about where those cars are piled up.  Authorities were asking family members of the missing to provide DNA samples.  I fear this is going to prove to be a disaster of historic proportions when the death toll is finalised.  

 

John 

[Ozexpatriate]

The brake practice when standing on a gradient does seem to be rather hit & miss to say the least - as do several other things that we do know about such as the apparent fire on the loco at Nantes (which seems to have been a put-it-out, walkaway, and leave it kind of job - without seemingly calling railway attention to the fire or if such attention was called then without the company doing anything about it (sorry, more questions than answers there).

This seems to be the most complete story so far about how much the railroad was informed of the shut down.

 

Nantes Fire Chief Patrick Lambert told the media that firefighters responded around 11:30 p.m. and put out the blaze, switching off the engine in the process, a move the company blames for releasing the brakes holding the train in place. They left about 45 minutes later, media has reported.

 

The Star has been unable to confirm this with fire officials.

 

Burkhardt said a company “track man” was on his way to the incident when he ran into firefighters leaving.

 

“This is where it becomes a bit inconclusive,” Burkhardt said.

 

He said the “track man” and the fire department eventually reported the incident to a dispatcher in Farnham, who should have called the two engineers staying in Lac-Mégantic — Harding and another due to pick up the train early Saturday morning. But he doesn’t know whether that happened.

 

“It was too late in any case. This train ran away, it was probably about an hour after they turned it off,” Burkhardt said. “Assuming (the dispatcher) had realized what the consequence would be, he wouldn’t have had time to rouse the two engineers . . . There wouldn’t have been enough time.”

As is always the case with such disasters, there are multiple failure points, often involving miscommunication. Edited July 9, 2013 by Ozexpatriate

[Ozexpatriate]

I fear this is going to prove to be a disaster of historic proportions when the death toll is finalised.

Even before we fully understand the loss of life, as a railway accident, the collateral damage to the town represents a railway disaster of historic proportions. (Most railway accidents don't extend beyond the perimeter of the railway infrastructure, though there are many that involved a much greater loss of life - usually passenger lives.)

[The Stationmaster]

A portable derail (if used) would likely have prevented this by derailing the lead loco as it started to move (think temporary catch point). They aren't 100% effective, though.

 

Adrian

I wouldn't argue with your final comment Adrian - having looked at pics on the 'net they remind me of some very similar devices I've seen tried over here and I'm sorry to say that from what I've seen of them if you hit them hard enough with sufficient axleweight or speed you might just as well have used a chocolate teapot for the job.  We used to have a lot of problems with broken 'flip over' steel scotch blocks at a depot where I worked (because folk forgot to flip them over) and things like some of these derails were tried instead - they got broken too.  

 

Basically there's no substitute for brakes, and if you want to stop something leaving a yard and safely(ish) derail it the only thing that is likely to work is a proper trap point - even with wimpy little British locos and rolling stock.

[stewartingram]

I'm familiar with the prodecures of investigattion in the UK, but can anyone explain (in simple terms) what the procedures will be over there for us please?

 

Stewart

[Fat Controller]

Even before we fully understand the loss of life, as a railway accident, the collateral damage to the town represents a railway disaster of historic proportions. (Most railway accidents don't extend beyond the perimeter of the railway infrastructure, though there are many that involved a much greater loss of life - usually passenger lives.)

The nearest I can think of is one four years ago at Viareggio in Italy:- http://www.telegraph.co.uk/news/worldnews/europe/italy/5696031/Italy-train-explosion-15-dead-in-gas-blast-after-derailment.html- though this involved LPG, and only two wagons exploded. What it might have been like had the whole train gone up, I wouldn't like to think- that area is very densely populated.

There was one in Belgium very recently, but fortunately there was only one fatality, due to asphyxiation- http://x.dawn.com/2013/05/06/belgian-chemical-train-accident-toll-rises-to-one-dead-49-injured/ Again, given the density of population, it was very fortunate that casualties weren't much higher.

[Ozexpatriate]

The nearest I can think of is one four years ago at Viareggio in Italy ...

 

There was one in Belgium very recently, ...

In terms of loss of life, the Saint-Michel-de-Maurienne troop train derailment of 1917 was second only to the Sri Lankan train caught up in the 2004 tsunami.

 

Wikipedia has a list of rail accidents sorted by death toll.

 

This (the Yaoundé train explosion in 1998) was a nasty rail tanker accident with a fuel fire.

Edited July 9, 2013 by Ozexpatriate

[beast66606]

Local reports now 13 dead and 50 missing.

 

It appears (again from local reports) as if the loco was shut down by the fire department when dealing with the minor fire, this released the brakes - allegedly the fire department failed to contact an engineer for clarification.

[Fat Controller]

In terms of loss of life, the Saint-Michel-de-Maurienne troop train derailment of 1917 was second only to the Sri Lankan train caught up in the 2004 tsunami.

 

Wikipedia has a list of rail accidents sorted by death toll.

The St Michel-de-Maurienne one was another runaway, though this time with a driver on board; he had been ordered, at gunpoint (Modane was under military jurisdiction at the time) to take the train down a 1/38 incline with very few functioning brakes. It narrowly missed being even worse- a Scottish regiment, bound for the Balkans, were on a train travelling in the opposite direction, which stopped just short.

[Adrian Wintle]

I'm familiar with the prodecures of investigattion in the UK, but can anyone explain (in simple terms) what the procedures will be over there for us please?

 

Stewart

 

If you look at the reports on the TSB website, it should give an idea of the processes followed:

http://www.tsb.gc.ca/eng/rapports-reports/rail/index.asp

 

This one in particular involved a loss of life (Aldershot VIA derailment)

http://www.tsb.gc.ca/eng/rapports-reports/rail/2012/R12T0038/R12T0038.asp

 

The Lac-Megantic one will be complicated since it is being treated as a crime scene, so there are multiple investigative jurisdictions involved.

 

Adrian

Edited July 9, 2013 by Adrian Wintle

[highpeak]

This site http://http://www.sdrm.org/faqs/brakes.html may help clarify the operation of air brakes. Pan Am's rule book on brakes is at http://http://www.guilfordrail.com/Bulls/Bulls2/Timetable%20and%20Rule%20Book%20Docs/PAR%20ABTH%20Rules%20121104.pdf Section 403 on page 100 details the procedure for securing trains to be left unattended.

 

An apparently reliable local source states that there were 4 propane tankers in the yard in Lac Megantic http://http://www.pressherald.com/news/propane-suspected-in-explosions_2013-07-09.html

 

MM&A has a less than stellar accident record, significantly worse than the industry average: "Federal Railroad Administration statistics show that in 2011, MM&A had a train accident rate of 10 accidents per million train miles throughout the company’s network, compared with a rate of 3.7 at Pan Am and a national average of 2.8 accidents per million train miles." (Bangor Daily News article, http://http://bangordailynews.com/2013/07/09/business/ten-years-of-highs-and-lows-for-montreal-maine-and-atlantic-railway/) Some of the accidents in recent years are not really the railroad's fault (a car that was being set out for repair ran away because of a fault with the brake rigging resulting in the brakes not being applied despite the actions of the conductor), others most certainly are, either various derailments resulting from poor track or a runaway trio of locomotives left with no handbrakes applied, air bled off, locomotives ran nearly a mile before colliding with a stationary train. The FRA reports make interesting reading.

 

This is going to be an interesting accident report and I hope it will be as exhaustive in scope as the analysis of the 1989 air crash at Dryden http://http://en.wikipedia.org/wiki/Air_Ontario_Flight_1363.The immediate cause of the accident appears fairly clear (air brakes were either inadvertently released during the firefighting operation or bled off as a result of shutting the unit down, insufficient handbrakes were applied to restrain the train once the air brakes were released, possible that previously applied handbrakes were released for some reason) but I think a lot of other questions about indirect factors need to be examined such as crew change procedures (is it wise to leave trains unattended for relatively long periods of time?); the relationship between poor track conditions, slow transit times and the impact on re-crewing; coordination with local fire departments (the fire chief, responding to the railroad's statement about the brakes being released by a third party, inadvertently revealed a flawed understanding of air brake systems) and the ability of a single engineer to cope with all the demands of this kind of operation, especially laying the train up at the end of his shift.

 

Beyond the accident investigation, the stakes are quite high for a number of railroads. For instance, Pan Am is running similar traffic in Maine, bound for the same refinery, over very marginal track. They have made some progress in upgrading their infrastructure, but such a long period of deferred maintenance can't be fixed overnight. Trains creep along at 10mph due to track conditions (and faster than 10mph and you run a serious risk of derailments due to harmonic rock and roll on stick rail with lousy cross-levels on staggered joints). They had a minor derailment in Veazie last week, the fact that the tank cars involved were either carrying CO2 or fumes will likely get lost in the debate that I am sure is going to start once the tragic mess in Lac Megantic is cleaned up.

 

Roads like MM&A and Pan Am can get away with lousy infrastructure and dodgy motive power when they are carrying newsprint, the stakes go up considerably when it's big oil trains. It will be interesting to see the politics of this play out.