Derailment and fire in Quebec

[The Stationmaster]

But even so, it doesn't exactly seem wise to hold an unattended train on a gradient, especially when it's carrying a dangerous cargo and there's a town at the bottom of the hill, without some way of preventing a runaway reaching the town. Even with more than enough handbrakes on, something could go wrong, including sabotage as there's no one around to stop it.

Personally I would describe it as downright stupid - all sorts of things could happen from a malicious act through to poor operating performance.  We learnt in Britain not to do it way back in the 19th century and haven't seen any reason to change since then!

 

Incidentally I find all the talk about charts of how many brakes to apply rather frightening and very different from the situation in Britain where the far more logical Instruction of 'apply sufficient hand brakes to hold the train' was used.  By not quoting a number but instead indicating what is required you do not mislead or inject false hop (or potential shortcuts), the brake force needs to vary to take account of rail conditions, weather conditions, rollability of the wagons (nowadays of course generally consistent with roller bearings), the condition of the handbrakes themselves, and the steepness of the gradient let alone the obvious factor of the weight you are trying to hold.  By quoting a number you automatically encourage folk to work to that number irrespective of the conditions, or taking account of the conditions, and you give them an excuse for sticking to it - an impractical and very dangerous way of writing an Instruction in my experience.

[BG John]

Personally I would describe it as downright stupid

So would I, but as you're an expert in such things, and my experience of holding anything on a gradient is limited to a van and trailer weighting less that 3.5 tonnes with rubber tyres on a road, I thought I'll be a bit restrained in my comments!

[highpeak]

Miscellaneous trawlings: Irving, who were the consignees for the oil, issued a press release that as well as expressing the natural condolences made it clear that it was neither their oil nor their train http://www.irvingoil.com/newsroom/news_releases/declaration_sure_la_tragedies_de_lac_megantic/

 

I didn't see this detail anywhere else but on this site http://www.liveleak.com/view?i=be2_1373307541&comments=1 the Nantes fire chief is quoted thus "Nantes Fire Chief Patrick Lambert said this was the fourth time in nine years that Nantes firefighters have put out fires aboard MMA trains." I think this was part of his response to Ed Burkhardt's suggestion that somehow the fire department were responsible for the runaway when they shut the engine down."

Edited July 13, 2013 by highpeak

[dave1905]

OMG.

 

Y'all just don't realize the scope of what you are talking about.

 

Saying that you would prohibit stopping a train carrying hazmat on a grade sounds really cool, but do you realize you are saying that you would prohibit stopping a train in an area twice the size of the UK? We have many lines where you are on a continuous grade for 150 miles. I just looked at one timetable for a portion of California and of the first 10 subdivisions, each between 110 and 150 miles long, over half had a continuous grade for the entire sub and the rest had over 50% of the sub on a grade.

 

I don't know how much hazmat you carry in the UK but in the US, pretty much an general freight train will have some hazmat of some kind in it. In certain corridors, over half of a general freight train could be chemicals. That doesn't count all the unit trains of chemicals.

 

US railroads have a 50 page or so booklet on how they are to handle hazmat already.

 

As far as the securement charts. We used to use the "sufficient" criteria and then when we got tired of crews guessing wrong and cars rolling out of yards and sidings, we had people put actual science to it and calculate, based on tonnage and grade, how many hand brakes were required. Surprise! When the crews follow the charts, cars don't roll out. So if you prefer to let the crews guess on what sufficient means on your railroad, go for it. We have the data that giving them a chart with science behind it is safer. We'll stick with the charts.

[highpeak]

I think the surprise expressed at the location the train was left in is more to do with there not being an additional layer of protection in place at Nantes to prevent a runaway beyond the train's braking systems. Bitter experience in the UK (e.g., the Stairfoot accident in 1870 http://en.wikipedia.org/wiki/Stairfoot_rail_accident) led to the provision of various means of preventing runaways beyond relying on just the brakes.

 

I agree with you that simply saying you can only leave a train on the flat more or less restricts railroading to parts of the Midwest and isn't workable. But I think there are things that can be done that would not be ruinous for the railroads and would provide a greater safety margin. As an example I found this railfan page http://www.coloradorailfan.com/sidings/sidings.asp?sub=1&s=1. When UP started using a passing siding outside Denver to park trains when the destination yard couldn't accept them, they installed a derail as a protection.

 

I think MM&A would have avoided a world of hurt if they'd thought a bit about what they were doing and done what UP did. (And yes, I know they didn't use the siding, they left the train on the main. The principle still stands. Use whatever track is most efficient,  put the protection where you are going to be parking your train.)

 

The issue with the idea of guidelines is, I think, the potential for crew to think that following the guidelines would mean they didn't need to then carry out a test of effectiveness. I can see where having a chart to follow saves you time guessing and carrying out repeated tests till you find the right number, but it would need to be stressed that the guideline is just that. And in fact the CN chart does more or less say that.

 

The CN rule book provided two charts to guide crews in setting handbrakes. One of them was intended for use when the train was being left on a grade. In the 2012 Hanlon accident the conductor followed the wrong guideline because he didn't realise the cut was being left on a grade. The chart he used suggested applying the handbrake on 2 cars (he actually only applied the brake on one car), the chart he should have used said 40% of the cars on a 1% grade (which was actually what the post-accident testing revealed was needed).

 

The crew did attempt an effectiveness test, but did it by trying to shove the cut backwards up the grade. The investigation felt this was not a valid test because it didn't really put the weight of the cars on the brakes.

[BR60103]

There seems to be a great reluctance on the part of regulators to actually force railways to do certain things.  In Red for Danger, Rolt mentions the perpetual refrain in inquiries of "Block, lock, and brake" which was delayed as long as possible. It was in my lifetime (well, I am a bit older than BR) that unbraked waggons were eliminated in Britain (they have been, haven't they?) 

 

Please remember that all the operations at the top of the hill were performed by one man -- engineer, conductor, brakeman and signalman if he'd had to put the train in the siding.

[kevinlms]

I think the surprise expressed at the location the train was left in is more to do with there not being an additional layer of protection in place at Nantes to prevent a runaway beyond the train's braking systems. Bitter experience in the UK (e.g., the Stairfoot accident in 1870 http://en.wikipedia.org/wiki/Stairfoot_rail_accident) led to the provision of various means of preventing runaways beyond relying on just the brakes.

 

But in that example NO BRAKES were applied and a single means (one sprag) of preventing a runaway were employed. If a minimum number of hand brakes were employed (plus one extra for additional assurance), may have meant that the accident would never have occurred. The fact is that a single means did actually stop the train from running away, until the additional wagons were shunted against, caused the failure. There is of course no way of telling, exactly how hard it was hit.

[highpeak]

But the lesson learned was that you need to build safety into the track layout.

[Gerald Henriksen]

I think you're right and I like your style. I am getting the sense that some rail accidents are viewed almost as inevitable in Canada. Three significant freight derailments in four weeks

 

 I guess it depends on how you define accidents.

 

Unfortunately, when talking about level crossing accidents, it is inevitable in North America because despite everything that is done drivers of cars and trucks still insist on getting in the way of trains.  There are simply too many crossings to eliminate, so the accidents will continue.

 

In terms of things like derailments, and other things entirely within the rail industries control, I think it depends on who you talk to.  I don't know that the TSB/NTSB feel that they are acceptable, but to a certain extent the simple numbers mean that stuff will happen.  There are over 2 million freight cars in use in North America, and the percentages simply at that point add up into real numbers even when you make the percentage into a really small number.  The number people in charge however ...

 

(all through different causes, its true, but the public just sees 'unsafe railways'). Look back at the TSB reports and you'll find two derailments of unit coal trains (2011 IIRC) six weeks apart because of broken wheels - wheels that were in an appalling state and yet one set was within acceptable standards).

 

If I found the ones you were thinking about they happened 11 months apart (R11V0254 and R11V0039), the indication is that both were within acceptable standards at their last inspection.  In the one case the wheel wasn't when failed, but the TSB indicated that an update to the AAR standard may have been to blame for allowing the wheel to continue in service.

 

The concept of putting voice recorders and cameras in locomotives, following the Burlington accident report, is based entirely on the assumption that there will be more such accidents and the only way to find out what happened is to know what the crew said and did. But that will be different each time and each accident.

So does that mean it is terrible that the authorities require cockpit voice recorders for planes? I mean doesn't that seem to imply that planes will continue to crash, and the only way to find out what happens is to know what the crew says?

 

We humans are not perfect, we all make mistakes but are fortunate that it most cases those mistakes are not life threatening.

 

So as long as humans are in control of trains, accidents will continue to happen. Having the most available information will allow the safety people to continue to make the system safer by removing any ambiguity in the causes of any future accidents. 

 

It's the system that needs to be looked at. In that case the signals should tell the engineer WHERE he's going, NOT how fast to go.

The system works when the engineer obeys the signal. If the engineer ignores the signal, then it doesn't matter how much or little information the signal is providing.

 

 

Of course even your preferred simple signal systems that don't convey speed don't work perfectly. Oops - http://www.standard.co.uk/news/transport/derailed-train-driven-too-fast-7442048.html

 

In other words, if your engineer is ignoring a signal, they can also ignore whatever other method you are using to indicate speed limits.

 

 

The Lac Megantic accident seems to have been a result of numerous incidents, all comparatively minor, combining together to create a catastrophe. That's not unusual. I've read dozens of accident reports (British, US and Canadian) in 50 years of railway journalism and that is a common factor in many of them. But if you can reduce the risk of those minor incidents, you take links out of the chain of events and a major disaster is not so easily created.

 

True enough, but you also have to carefully look at your list and decide which really did contribute to the incident, and those that reflect a personal judgement on what things should be.

 

Locomotives running round in patched up 30year old paint (on a railway that likes its red and gold) suggest to me that they aren't necessarily well maintained.

 

So does that mean the class I's are maintaining their locomotives? After all CP ran a patched locomotive for 6 years (see SOO 4513 http://www.canadianrailwayobservations.com/croarchives/feb2012cpaaa.htm ).  And I am sure if you search the photo sites you can find lots of class I locomotives still using 20+ year old faded paint jobs as well as other patch jobs.

 

Painting a locomotive is expensive, both in money and time out of service.  It is often a better option to simply patch the reporting marks to reflect the new ownership, and this has nothing to do with how the locomotive is being maintained.

 

Similarly, MMA was using leased locomotives - certainly the train that was parked east of Lac-Magentic after the crash had several leased locomotives in the consist.  But again, the class I railways also lease locomotives, so this doesn't indicate anything either.

 

Which as I said before doesn't say that MMA was doing things properly, but rather the paint job does not tell you anything about what was or was not being done.  There is a saying that is appropriate here - don't judge a book by its cover.

 

 

In any event five old locomotives instead of one or two new(er) ones must increase your chances of trouble.

 

Again, you are seriously overestimating the savings.  Those 5 older locomotives aren't all there in case of failures, they are there because that is the amount of power needed to move the train.  Thus you need 3 new locomotives to do the job of those 5 older ones.  With only 1 or 2 newer ones it is likely that the train isn't even going to be able to leave the yard.

 

Even with all the new locomotives that have been bought over the last decade railways like CP are overhauling/upgrading existing units, or sending them off to be rebuilt (see SD30C-ECO and GP20C-ECO) because that is cheaper than paying the cost of new units.

 

But as I believe someone else posted, really the engine is not what caused this.  If they were relying on the engine to be running to maintain the air brakes to keep the train stationary, then MMA were doing things wrong.

 

More to the point though, little or nothing seems to have been done to make this railway better able to cope with this significant extra tonnage, which is likely a revenue windfall. Where were the track upgrades?

 

Nothing has been reported, so we don't know if MMA were upgrading track or not.  It is even possible that (as shown in the graph of the article you linked to) given the surge in oil shipments started 18 months ago MMA (like any other company) may or may not have had time to actually do any improvements bearing in mind that winter around here wipes out a lot of your available time.

 

But in this case the track didn't cause the derailment, so it is a non-issue.  The derailment (likely) occurred where it did because of a 10mph switch, which short of removing that yard there likely wasn't anything that could be done.  And even if you could upgrade the switch to say 30mph, the train was still too fast.

 

Sure, trap points have been phased out on passenger lines in the UK but this is a freight line in Canada where they could do their job just fine. Furthermore, you can forget to set and lock a derail. You can't reset for the main line without setting the trap point.

 

Good point, being interesting to see if the TSB says anything, or for that matter if they have said anything in the past.

 

Yes, Canada has trouble with kids getting onto property and spraying graffiti, so you NEED to have secure places where hazardous goods trains can be stabled up in relative safety - yes, decent fences, gates, lighting, and security guards if necessary - it's only going to be one or two carefully planned places on the route.

 

My point was that a lot of this graffiti is being done in large city yards, that already are fenced, gated, security guards, etc.  So if those measures aren't working where they are implemented, what will extending them to the middle of nowhere do?

 

And do you really think a security guard at night in the middle of nowhere is going to to their job night all night?  Isn't more likely they will fall asleep, talk to a girlfriend, play games on their phone, etc.?

 

We live in dangerous times and Canada has two alleged railway terrorists in custody.

 

So we also have 2 alleged Canada Day terrorists in custody - http://www.cbc.ca/news/canada/british-columbia/story/2013/07/02/bc-rcmp-terrorism.html - does that mean we should fence off and create secure spaces for all future Canada Day celebrations?

 

And of course I assume that the London Undergound is searching everyone who enters their stations given their terrorist history.

 

I don't mean to trivialize the terrorist issue, but at a certain point you have to rely on the police to deal with the issue.  You simply cannot lockdown entire countries.

 

 

Is it reasonable to expect one man to drive a complete shift at mind-numbingly low speed, hour after hour, to tie up his train at dead of night, to walk 1,000yds climbing up and down on 76 cars to screw down handbrakes and then 1,000yds back again (sometimes in VERY inclement weather) - a job which will take another hour or more?

 

This I fully agree with, and it will be very interesting to see what the TSB says about this.

 

There's a lot in here that has far more to do with wringing the last ounces of profit out of a tired old railway, than with the safe efficient 21st century movement of dangerous goods.

 

Well, I suspect that profit and the MMA aren't two things that really go together very often.  This article mentions yearly losses of $4.5 million - http://www.montrealgazette.com/news/Railway+companies+scaling+down+single+crew+member+some+lines/8632006/story.html

 

Finally, there's the PR aspect of not having an incident plan that says who talks to the press and what they say. So you add a PR disaster by having your Chief Exec first blame the Fire Brigade, then a 'non-railway person' and finally rounding on his own engineer and hanging him out to dry before he's even been charged with anything, never mind tried. It just might be that something good can come out of this tragedy if it is seen as a wake-up call.

We don't know if MMA had an incident plan or not, though I would tend to not.  But if your CEO is going to make things up as he goes, and incident plan wouldn't help anyway.

 

 

Just like the railway industry in the UK has its issues, the industry here does.  

 

Because there are no massive government subsidies to keep the system running, every expense needs to be justified.  That means some things you may view as important - new paint jobs - just aren't done over here.  Keeping the trains running is more important than paint (and the freight doesn't care).

 

Because the railways pay property taxes on their rail lines, there is a monetary incentive to keep things as simple as possible, even if that means things are less than ideal.

 

Because the system is low density (compared to Europe), things that are taken as a given elsewhere don't work.  It's not cost effective to signal everything, and it is impossible to secure more than an very small fraction of the system.

 

Which is not to say the system is perfect (it most certainly isn't), but that it is different.

 

Can it be improved?  Yes.

 

But it is also worth remembering that as bad as you may feel rail is in North America, it is still safer than trucking.  And if the oil doesn't go by rail, and can't go by pipeline, then it will go by road.  We aren't willing to give up oil as an energy source.

[kevinlms]

But the lesson learned was that you need to build safety into the track layout.

But I read that the prime safety aspect is to ensure that any part of the train DOESN'T MOVE. It is a 2nd back up to suggest that if vehicles do start to roll then they should be derailed by some means.

 

Here is a similar example where the wagons were left on the mainline (not against any rules), with an almost identical series of events to that of Stairfoot.

 

http://en.wikipedia.org/wiki/Abergele_rail_disaster

 

The main cause being the dodgy shunting practice employed by the stationmaster (not RMWeb's!) where as it was being done in front of a due train. Almost certainly the job was being rushed to avoid delay to the Irish Mail - the most important train on the line. Tragically, some of the wagons carried paraffin oil, the resultant fire inflicting most of the deaths, in the wooden coaches.

[Gerald Henriksen]

Incidentally I find all the talk about charts of how many brakes to apply rather frightening and very different from the situation in Britain where the far more logical Instruction of 'apply sufficient hand brakes to hold the train' was used.  By not quoting a number but instead indicating what is required you do not mislead or inject false hop (or potential shortcuts), the brake force needs to vary to take account of rail conditions, weather conditions, rollability of the wagons (nowadays of course generally consistent with roller bearings), the condition of the handbrakes themselves, and the steepness of the gradient let alone the obvious factor of the weight you are trying to hold.  By quoting a number you automatically encourage folk to work to that number irrespective of the conditions, or taking account of the conditions, and you give them an excuse for sticking to it - an impractical and very dangerous way of writing an Instruction in my experience.

 

I suspect that it is also a lot easier in Britain with your lighter and shorter trains.  Trial and error on a 10 car freight is easy, not so easy when you start talking 70, 120, 200, etc. cars.

[dibber25]

This has certainly been an interesting discussion. As an outsider looking in, I see things which alarm me  and I also see things which are the way they are because North America is very different from the UK. I also feel that the tit-for-tat arguments that 'you have accidents in the UK, too' are not relevant. Of course, no system is perfect and of course we have accidents. Many of ours, too, are often these days related to bad behaviour by motorists, but that is irrelevant in this case (none of the freight accidents to which I referred involved road vehicles). We also have a pretty good safety record and, I would say, an exceptional record for learning from our mistakes and changing and upgrading our systems accordingly. The installation of TPWS to override the effects of a SPAD being just one example. But I see standards and practices which are acceptable in North America and which I don't believe should be. I also feel that there's an attitude - even evident in some of the posts - that 'we do things our own way, we've loads of experience, mostly we get it right, and we're quite happy that we don't need to make any changes or learn from anyone else'. No safety-critical changes were even suggested in the Burlington report and I suspect that none will be after this, either. Just a tightening of the rules to try and ensure that people do what they are supposed to. To me, that means that Canadian rail safety is stuck in a rut, with TSB reports saying far more about the psychology of why human beings make mistakes, than about the technology to defend against those mistakes. I'm sorry, but I am utterly convinced that traffic increases of the kind involved here require investment in improvements to infrastructure and equipment, maybe other aspects as well - who pays, is a different issue (government, railroads, oil companies - but someone needs to). Otherwise, perhaps it IS best that investment goes into pipelines, (although that's not something I actually believe) and that railways which prove unable, for whatever reason, to carry traffic safely, should close.

CHRIS LEIGH

[Grovenor]

dibber25, on 13 Jul 2013 - 18:55, said:

Sure, trap points have been phased out on passenger lines in the UK but this is a freight line in Canada where they could do their job just fine. Furthermore, you can forget to set and lock a derail. You can't reset for the main line without setting the trap point.

 

 

There is no difference in this respect between a trap point and a derail. Either can be left independent of the main line switch or rodded to it to work together, and it is the provision or not of the linkage not the type of device that makes the difference. In a situation like this it is unlikely rodding would be provided in either case as it would make it very difficult for the crew to operate the switch, especially under a few feet of snow and ice.

Edited July 14, 2013 by Grovenor

[The Stationmaster]

OMG.

 

Y'all just don't realize the scope of what you are talking about.

 

Saying that you would prohibit stopping a train carrying hazmat on a grade sounds really cool, but do you realize you are saying that you would prohibit stopping a train in an area twice the size of the UK? We have many lines where you are on a continuous grade for 150 miles. I just looked at one timetable for a portion of California and of the first 10 subdivisions, each between 110 and 150 miles long, over half had a continuous grade for the entire sub and the rest had over 50% of the sub on a grade.

 

I don't know how much hazmat you carry in the UK but in the US, pretty much an general freight train will have some hazmat of some kind in it. In certain corridors, over half of a general freight train could be chemicals. That doesn't count all the unit trains of chemicals.

 

US railroads have a 50 page or so booklet on how they are to handle hazmat already.

 

As far as the securement charts. We used to use the "sufficient" criteria and then when we got tired of crews guessing wrong and cars rolling out of yards and sidings, we had people put actual science to it and calculate, based on tonnage and grade, how many hand brakes were required. Surprise! When the crews follow the charts, cars don't roll out. So if you prefer to let the crews guess on what sufficient means on your railroad, go for it. We have the data that giving them a chart with science behind it is safer. We'll stick with the charts.

Oh dear - seems we are divided by a common language.

 

Of course we can - and always have - stop trains on steep gradients and we run trains over them but the difference is we don't dump them there unless they are properly protected and secured and even then every effort was made to avoid building sidings on gradients.  But leaving a train unattended on a steep gradient, especially on a running line is - as I've said - downright stupid but if that's the way people want to play it don't cry when it all goes wrong and what is statistically likely to happen actually happens.  As I said we learnt our lessons back in the 19th century and took the necessary steps to try and avoid such things happening again (by odd coincidence the significant event which changed thinking involved wagonloads of oil and a resultant fire which claimed lives although the circumstances were slightly different).

 

As far as not trusting staff to do the job they are paid and taught to do we are again talking, I suppose, about a gulf of attitude and understanding which stems from years of practice.  In Britain - and most of Europe and the British colonies people were trusted to be competent to do the jobs they were paid to do.  Now obviously this sometimes went wrong so systems were changed and equipment altered but the principle of trust still held until relatively recent years when the safety mafia - generally folk of limited understanding of what was actually involved - came out with all sorts of daft and over-prescriptive ideas.  This often means seemingly treating the folk who are doing the job as semi-brainless and incapable of thinking for themselves but in reality means a system of back-covering for the management and insurers.  

 

If the relevant operating staff cannot tell whether or not a train, or vehicles, is properly secured by appropriate  testing or by simply looking at it then they shouldn't be in the job.   Telling them how many vehicles to brake is dangerous - it leaves a hole by creating a trap and anyone with on the ground experience can immediately see that although it is hardly surprising that the likes of NTSB working on a theoretical basis come out with such ideas.

[The Stationmaster]

I suspect that it is also a lot easier in Britain with your lighter and shorter trains.  Trial and error on a 10 car freight is easy, not so easy when you start talking 70, 120, 200, etc. cars.

Yes & no is the simple answer to that one I think.  Human nature tends to take the shortest and easiest (and fewest number of handbrakes) route but in terms of trains size it all becomes a matter of what you are used to and the number you have to brake in any particular location becomes established by experience.  But if you test the efficacy of the applied handbrakes it quickly becomes apparent if you've done it right - and the tendency then might even be to apply 'an extra couple for luck' to avoid not having enough and having to go back and do it again.

 

So in this incident it could well be that the single manning/hours on duty/tiredness etc starts to assume considerable significance as facts emerge.

[ZiderHead]

So in this incident it could well be that the single manning/hours on duty/tiredness etc starts to assume considerable significance as facts emerge.

 

Certainly looks like it. I wonder how long it takes one guy to apply say 40/100 handbrakes on a mile-long train in the middle of the night on an unlit section of track after a long, boring shift? Especially if the train is parked on an uphill gradient (assuming the brakes are applied on the downhill end of the train) and he has to walk the entire length of the train. Not a 10 minute job I imagine.

 

When you are dealing with such long trains, especially single-manned, there seems to be a good case for a simple system to apply all the handbrakes using electric motors, although that does introduce a potential electrical hazard for flammable loads ...

Edited July 14, 2013 by ZiderHead

[davknigh]

All this discussion about the safe storage of trains on sidings is good but it all boils down to this, for the price of two extra part time staff the train could have been kept moving and this discussion wouldn't be happening at all. I think 'false economy' sums it up nicely.

[Edwin_m]

It was in my lifetime (well, I am a bit older than BR) that unbraked waggons were eliminated in Britain (they have been, haven't they?)

They stopped running routinely in around 1990 and new unbraked wagons were being built up until about 1960 - no doubt somebody can give better dates.  Something that should have happened much sooner in my view, and where North America and practically everywhere else was well ahead of the UK. 

 

There were many serious accidents with unbraked wagons over the years, and others involving other trains being derailed by the sprung catch points that had to be installed on main line gradients to derail any portion of an unfitted freight that split and rolled back before it could do more serious damage. 

[dave1905]

To actually "solve" a problem you have to get to the root cause.  If the problem is how to stop a train from rolling away, then that's what you solve. 

 

Just saying don't stop the train on a grade isn't a solution because its impractical.  If a rock the size of a school bus falls off a mountain and lands on the track, its going to take a day or two to remove it (and yes this happens several times a year in NA, the hazards of operating across mountain ranges).  Trains will stop, some will be on grades and no they won't be manned while the rock is jackhammered and blasted out of the way.  We are also going to have lines closed by blizzards, ice storms and avalanches.  Its a hazard of where we operate.  Trains are going to stop and be unattended on grades.  Saying don't do it just creates a huge safety gap because when it does happen there isn't a way to deal with it effectively.

 

Every time any discussion about railroad safety comes up people from the east side of the pond always want to add derails (regardless of whether the incident they are discussing could have been prevented by a derail).  The thing about a derail is it creates something bad when it works.  Derailing a train doesn't eliminate risk, it just makes a smaller pile.  The goal is accomplish the task with creating any pile.  Even if you had a derail you would still need the exact same processes for securement as you did without the securement.

 

I am in no way endorsing whatever the MMA did or its operating practices.  The whole concept of the spun off shortline was that the big railroads didn't find those routes profitable. In order to operate them the shortline is going to have to reduce costs.  Typical methods are reduced crew sizes, reduced crew pay, leased locomotives (cheap leased locomotives), reduced main track speed (reduced maintenance costs).  

 

The major US railroads spend billions of dollars in new rail, new ties, new infrastructure, better communications, better safety processes.  There are some risks we aren't able to engineer out yet (there are parts of the rail an ultrasonic rail defect detector can't scan, so defects there can't be detected).  There are also things that we've been doing for decades because we thought they were safe that we've found really didn't increase safety at all.  Giving the crews a securement chart isn't saying that you don't trust the crews, any more than giving a mechanic a chart telling him what torque to apply to bolts when assembling an engine is saying you don't trust the mechanic.  Its giving them a tool, the railroad still has to trust they will use the tool to do the job right.  By the way we also have pages and pages of instructions on how to build trains operating on a grade, how to place and ship hazardous materials, systems to detect errors and processes to correct those errors, plus processes figure out what caused them.

 

Shortlines lack a lot of that infrastructure.

[Fat Controller]

They stopped running routinely in around 1990 and new unbraked wagons were being built up until about 1960 - no doubt somebody can give better dates.  Something that should have happened much sooner in my view, and where North America and practically everywhere else was well ahead of the UK. 

 

There were many serious accidents with unbraked wagons over the years, and others involving other trains being derailed by the sprung catch points that had to be installed on main line gradients to derail any portion of an unfitted freight that split and rolled back before it could do more serious damage. 

The last to be built were in 1971, IIRC, for alumina traffic between Blyth and Lynemouth. This was a captive fleet that stayed on this circuit of freight-only routes until their replacement by air-braked stock at the beginning of the last decade. On continental Europe, some grain hoppers were built in France without even hand brakes as late as the 1960s!

There was a derailment attributable to catch points in the UK within the last decade, IIRC, when a fully-braked train carrying petroleum products stalled on a gradient in NE London. The train started to roll back, possibly only by the amount of slack in the couplings, and the catch points did what they were meant to.

[The Stationmaster]

The critical thing about a derail or a trap point - especially the latter is that it gives you a choice of where the errant train/vehicle will land.  Yes it can still lead to big pile up (I think the highest heap I ever saw was about 20 feet high) but you make the choice of where it will happen if it is going to happen at all.

 

Trap points were - as we've already mentioned - done away with on passenger lines from sometime in the 1980s because it was considered that the risk and consequences of a train running over the trap were statistically far worse that the risk and consequences of a sidelong collision where no trap existed.  But at much the same time signal overlaps tended - wherever possible - to be increased to a full overlap distance instead of those cases (not all) where the trap was only a short distance in advance of the signal; adding TPWS has of course further mitigated the risks of a collision.  On freight lines traps are still used - in fact in a scheme which I devised in the early 1990s I had put in to enable us to get the required distance between signals in order to stand coal trains to run-round and the line was taken out of passenger use as a consequence.  And yes derailments - usually straight run throughs - have occurred at trap points.

 

Derailments on catch points are a rather different kettle of fish and tended to be associated with roll backs, often due to couplings slacking out,  or locos getting overpowered and trains stalling or rolling back, plus of course the catch points working as intended in the event of breakaways and preventing the errant vehicles from going anywhere more harmful.  But as we went over to fully fitted freights and - more importantly - fully fitted engineering trains so catch points were removed.  The conversion to fully fitted began in the first half of the 1980s at the southern (N.B. not Southern) end of the country and spread northwards mainly depending on the speed at which wagons for engineering trains could be converted.  By 1992 there were a handful of workings left using unfitted wagons in particular special circumstances and under very strict controls with all of them effectively against a timescale to end - they were all running in areas which were otherwise fully fitted.

 

Incidentally working in South Wales for some years I was well aware of the differences that impacted on trying to hold a train on handbrakes - you simply could not legislate any sort of number because what worked on a dry sunny day would be totally inadequate on a wet one or it might even be inadequate on a dry sunny day if a train was loaded with wet coal off the ground.  Similarly on some wagon types the handbrakes were notoriously poor in practice whatever the theories might say so, again, from experience, the folk doing the jobs knew when and what with they had to take special care - and we staffed to do the job safely.

[burgundy]

My apologies if I have missed the point somewhere in the preceding 11 pages, but why did the train have to be parked at all? If it was a regular stopping point, why was there not a second driver waiting to take the train on its way, while the first one went to rest? Stationary trains don't make money - especially when they have to burn fuel just to stay still!

Best wishes

Eric

[Fat Controller]

My apologies if I have missed the point somewhere in the preceding 11 pages, but why did the train have to be parked at all? If it was a regular stopping point, why was there not a second driver waiting to take the train on its way, while the first one went to rest? Stationary trains don't make money - especially when they have to burn fuel just to stay still!

Best wishes

Eric

It would seem that the regular change-over point for drivers was at the American- Canadian border; presumably the driver couldn't have made it within his allotted hours.

[Gerald Henriksen]

My apologies if I have missed the point somewhere in the preceding 11 pages, but why did the train have to be parked at all? If it was a regular stopping point, why was there not a second driver waiting to take the train on its way, while the first one went to rest? Stationary trains don't make money - especially when they have to burn fuel just to stay still!

Best wishes

Eric

 

I don't believe it has been discussed, but the simple reason is there was no crew to take the train further.  It appears, given that there was another train east of Lac-Megantic, that the opposing trains arrived at Lac-Megantic and parked, swapping crews the following day.

 

As for the "stationary trains don't make money", while true it can also be true that a stationary train loses less money than a moving train.  If you want to keep both trains moving, you need to double the number of employees, and employees are one of the major costs of running a business.  Presumably somebody did the numbers and it was cheaper to leave 2 locomotives running overnight than to hire 2 more crew.

Edited July 14, 2013 by Gerald Henriksen

[highpeak]

I quoted a thread off the trainorder website that gave some information about MM&A's working practices. The territory for what they call Job 1 was from Farnham (43 miles east of Montreal) to Lac-Megantic. Job 2 is from Brownville Jct to Lac-Megantic. The operator for Job 2 was taking his rest period. It is not clear whether his train was parked already at Nantes (I have seen a reference to there already being a train in the siding at Nantes, but not clear if that was the case, if it was the train from Brownville Jct, a work train, no doubt that will emerge in the investigation as opposed to journalists/bloggers etct).

Lac-Megantic is not a great place to lay a train up idling. It was traditionally in CP days a crew change point, but crew changes are not necessarily going to happen in a well-coordinated manner any more because there are too many factors against that, some that MM&A could control (track condition, reliability of motive power, number of available crew), some they can't (connections with other railroads, hours of service rules, weather conditions and so on). Laying a train up in Megantic for several hours may have involved splitting it to avoid blocking grade crossings with all the implications of that, and certainly would have led to issues with the residents with a diesel idling for any length of time (google "idling trains PAR" to see how well that goes over).

There is no huge rush to get the train east. Until recently capacity issues at the refinery in St John had led to trains being occasionally laid up in a number of locations on Pan Am while they got the whole operation settled down to the right flow of crude. That seems to have been sorted out, I think Irving expanded either their reception facilities or the refinery itself.

So Nantes is the choice of crew change point for a number of reasons. (If you look at a map, you'll see there is really nothing beyond Megantic until you get to Jackman or Brownville Jct.

And while all of that is worth looking at and no doubt will be, the real issue is that the infrastructure at Nantes did not provide any redundancy in safety so that if the primary means of securing a train (brake systems) failed to do its job, for whatever reason, then there was no second layer of protection.