Oil Train

[Wizlish]

dehusman

Wizlish

Only if you're using the word 'impending' to describe catastrophic consequences. Anything an IR detector finds is likely already a massive, complete failure as regards bearing or even wheel integrity.

 

Well not really.  Modern IR hot bearing detectors are set up in the back office so the deteector tracks the temps of the individual axles.  If one axle is trending hotter over several detectors, it will alert and that car can be set out.  A significant number of "hotbox" alarms are false positives.

What I meant was 'by the time something like a bearing problem has become visible to an IR detector, it will have been suffering consequences of failure long enough to be running hot, and as you accurately note, probably in the context of progressing to heating up enough to trigger the detector during the time from the last wayside detector check.'   (EDIT -- I missed one of the points here, which (as will be subsequently discussed in more detail) the detector system notes increased heat-trace trends per axle, even when well below 'failure' threshold, and this is likely what both buslist and Dave were discussing.  Consider that point made as stated, and if we assume that the train is correctly stopped and an inspection made each time such a temperature trend is detected, my comments regarding that aspect of the IR system are incorrect.)

I don't want this part of the discussion to turn into the pros and cons of thermal detection of bearing failure, which isn't the topic at this point (even if it is one cause of possible derailments).  What I'm looking at is continuous detection of derailment 'signatures' and appropriate notification action (which, again, is NOT 'automatically putting the train in emergency every time a possible derailment state is detected).

 

And -- comparable to the situation at Marysville -- it's a crapshoot whether such a problem that has progressed to such an extent fails before or after it happens to pass by a wayside detector and generates enough warning to stop the train

 

Its my understanding that the recent derailment was an axle failure (not a journal failure).  There are no detectors for axle failure (it is an incredibly rare failure) and the detection is usually after failure when it hits the ground or puts the train in emergency.

 Again, 'comparable' only in the specific sense that a wayside installation -- here, the dragging-equipment detector tdmidget mentioned -- would not likely have helped the situation, unless it 'just so happened' that the broken equipment passed over it before the failure progressed to an accident.  However, an onboard detector would have caught this almost immediately.  I do think it is fair to say that it is 'comparable' to note this both for the Marysville incident (as supposedly detectible/preventable by dragging equipment detectors) and for other causes of derailment as they might be detected solely by wayside installations.

 

If I recall correctly this is one of the reasons for increasing the number of detectors, and I'm not questioning that idea in the least...

 

If you put a hot box and dragging equipment detector every mile (IR or acoustic) it wouldn't have detected an axle failure prior to failure (nor would it have detected a center beam failure, a side sheet failure, a draft gear failure or any one of a hundred other possible component failures).

Essentially my point; I really couldn't have stated it much better.  It may be interesting to see what buslist notes on that subject, as it may be that there is some 'sweet spot' in cost-effective wayside detector-suite spacing that will give "good enough" results short of the cost of an on-train system.  You have now raised additional criteria that, to me, advocate for an on-train system [EDIT -- strictly for those failures that, likely unlike most bearing problems in practice, can't be caught through 'progressive' tracking from a proper system of intermittent wayside detection], although I have to wonder how many of those  conditions aren't about as 'incredibly rare' unlikely as an axle failure...

• Reply

[dehusman]

Wizlish

Only if you're using the word 'impending' to describe catastrophic consequences. Anything an IR detector finds is likely already a massive, complete failure as regards bearing or even wheel integrity.

Well not really.  Modern IR hot bearing detectors are set up in the back office so the deteector tracks the temps of the individual axles.  If one axle is trending hotter over several detectors, it will alert and that car can be set out.  A significant number of "hotbox" alarms are false positives.

And -- comparable to the situation at Marysville -- it's a crapshoot whether such a problem that has progressed to such an extent fails before or after it happens to pass by a wayside detector and generates enough warning to stop the train.

Its my understanding that the recent derailment was an axle failure (not a journal failure).  There are no detectors for axle failure (it is an incredibly rare failure) and the detection is usually after failure when it hits the ground or puts the train in emergency.

If I recall correctly this is one of the reasons for increasing the number of detectors, and I'm not questioning that idea in the least...

If you put a hot box and dragging equipment detector every mile (IR or acoustic) it wouldn't have detected an axle failure prior to failure (nor would it have detected a center beam failure, a side sheet failure, a draft gear failure or any one of a hundred other possible component failures).

[Wizlish]

Buslist

The industry did a cost benefit on on-board detectors a few years ago and found far more problems than benefits.

Not to criticize in any way your professional competence as one of the better-qualified and knowledgeable contributors here -- You say this and don't provide a source, citation, or link?  Don't indicate what the 'problems' were, let alone with any sort of indication why they were greater than the equally unindicated) benefits cited?  Please document, and at least synopsize the detail conclusions here.  A discussion can't go forward just on use of older technology and unqualified appeals to authority.

So where does this detector get its power? How does it communicate to PTC? Via the locomotive? Fairly strong radio signal required and of course additional already scarce bandwidth.

None of this is technically 'showstopping' (as yet), although I really can't describe much of the technical detail without causing NDA problems.  There is at least one critical issue involving methods of wireless communication, but I understand there is European OTS (or at least designed) equipment that already purports to do what would be needed in the necessary respects.  (Yeah, I know, looks like crank science, but really isn't.)

Actually this would be inferior to IR detectors as IR detectors find impending failures ...

Only if you're using the word 'impending' to describe catastrophic consequences.  Anything an IR detector finds is likely already a massive, complete failure as regards bearing or even wheel integrity.  And -- comparable to the situation at Marysville -- it's a crapshoot whether such a problem that has progressed to such an extent fails before or after it happens to pass by a wayside detector and generates enough warning to stop the train.  If I recall correctly this is one of the reasons for increasing the number of detectors, and I'm not questioning that idea in the least...

... this detector would [not] find failures unless you are proposing sensors on all eight bearings. How many hot box setouts are there without a derailment? The vast majority I would bet.

Not really all that difficult, technically -- although remember we're talking about derailment detectors, not hotbox detectors, at this particular moment.  I for one prefer the idea of a device that detects bearing failure more directly, finds and alerts on broken axles with appropriate immediacy, and can do some correlation or sensor fusion with other sources of information.

This does presuppose something 'more' than the reported Spanish detectors (which, however, could be 'improved' at least imho by incorporating a few simple things like recognition of sideframe/bolster vertical angle and rate-of-change information on truck rotation and skew).

Even better are the acoustic [note sp. if not a typo] bearing detectors that find bearing defects long before they become impending failures.

Yes, but there apparently have been problems in isolating some of the acoustic components of the signal from other noise a wayside acoustic detector would pick up.  I wholeheartedly affirm that acoustic detectors ought to be placed at least as densely as IR detectors -- that is to say, ultimately at short enough intervals (at least where HHFT trains and their outsize public 'concern' run) to provide at least political assurance that a developing 'derailment cause' will be caught before it is likely to cause an actual derailment.  In fact I think it makes sense to have both IR and acoustic wayside detectors even in situations where some form of 'derailment detector' is provided on individual cars. 

The problem that Euclid has identified, to recap briefly, is that there really is, and I agree in some ways with him that there can't be, any way to make a HHRT tank car particularly physically safe in a derailment as things currently stand.  That means that the industry must concentrate on things that reduce the likelihood of derailments, even when the causes of derailments occur quickly or unexpectedly.  (It would be nice if derailments could be prevented with 'perfect' attention to equipment maintenance, wheel grinding, axle or wheel NDT (and why not?) and the like, but I think there is still a significant constellation of derailment causes that happen even if the train's equipment and the track pass their testing.  And I think those, too, need to be detected -- certainly while there is any chance of preventing an actual derailment.  If there is an actual derailment event, I think there would be clear advantage in passing knowledge of it to the crew as promptly as possible, even if there is no cost-effective way to assure 'automatic stop' in minimum time or distance in that particular instance.

The 'other half' of the problem with Bakken/Eagle Ford HHFT traffic is the undocumented volatile content.  I'm certainly no expert, but I think that mandatory degassing before shipment (or whatever that's properly called) will go a very long way to addressing the 'real' public problem, the incidence of fireballs and catastrophic sequential thermal breaching in oil-train accidents (even those that seem to be comparatively slight).  THAT is where the Federal agency people  and politicians ought to concentrate attention, particularly now that it seems railroading is losing much of the 'cash cow' traffic that would fund mandates like Feinberg's proposed ECP instantiation.

[schlimm]

Buslist

The industry did a cost benefit on on-board detectors a few years ago and found far more problems than benefits.

Likely that was before the transport of oil surged.  The avoidance of disastrous derailments of oil would change the outcome of that analysis now.  And the industry has a duty to run safely, even if that cuts into profits.  It's part of the cost of being in business responsibly.

[Euclid]

There is a derailment detector developed in Spain as I recall from earlier in this thread.  That one does not use electrical power.  It mechanically senses a derailed truck and dumps the air on the train.  There is one on each truck.  I believe the ideal derailment detector, however, would not make an emergency application as a response.

For U.S. application, dynamiting the brakes in response to a derailment on relatively long trains might be even more of a risk than in Spain.  I see the ideal application in the U.S. being on crude oil unit trains if the ECP brake mandate remains in place.  With ECP, the power can be brought to the sensors, and signals can be sent to the brakes for brake applications other than the emergency application.  It could focus the braking effect ideally on the point of derailment to prevent jackknifing.

Derailment detectors would not replace other wayside detectors that find defects.  They would be a backup in case a derailment happens without any predictive defect detected.  Indian Railways is applying derailment detectors to passenger trains.    

[Buslist]

Wizlish

 

As Euclid indicates, an accident like this is a 'poster child' for actual derailment detectors, probably devices mounted on individual cars that can signal promptly when problems are detected.  To some extent such a device can also check or track bearing problems better than wayside IR detectors.

Personally I think any of the money Sarah wants to spend on ECP ought to be put dollar for dollar into this technology instead, until all the HHFT and PIH cars have been instrumented and the added programming (to PTC) for alerting and information exchange with the devices has been done.  I'd even go so far as to state implementation of these devices is a priority over the boondoggle that "PTC" has become at this point -- it certainly offers a much more positive and direct form of 'safety' provision.

 

So where does this detector get its power? How does it communicate to PTC? Via the locomotive? Fairly strong radio signal required and of course additional already scarce bandwidth.

Actually this would be inferior to IR detectors as IR detectors find impending failures, this detector would find failures unless you are proposing sensors on all eight bearings. How many hot box setouts are there without a derailment? The vast majority I would bet. Even better are the acustic bearing detectors that find bearing defects long before they become impending failures.

The industry did a cost benefit on on-board detectors a few years ago and found far more problems than benefits.

[Euclid]

tdmidget

AKA dragging equipment detectors. Quite possibly what alerted this crew. How many times are you going to reinvent the wheel?

A dragging equipment detector would detect a derailment, but it is a wayside stationary detector, whereas a true derailment detector is mounted on the rolling stock.  It, of course is possible that a dragging equiment detector happened to alert the crew of the CSX train if there happened to be a dragging equipment detector between Middlesettlement Rd. and Old Mt. Tabor Road.

But derailment detectors on board are hardly reinventing the wheel, as you say.  A derailment sensor on each truck finds a derailment the instant it happens.  A dragging equipment detector has to wait for the derailment to arrive.

[Wizlish]

tdmidget

AKA dragging equipment detectors. Quite possibly what alerted this crew. How many times are you going to reinvent the wheel?

If dragging equipment detectors alerted this crew, they did a piss-poor inadequate job of it.  Are you suggesting that dragging equipment detectors of the usual types be placed every mile, or at the entrance to every curve, so an incipient HHFT derailment will be caught before it progresses to damage and fire?  And then discriminate between actual dragging equipment and broken axles or derailed trucks?  That technology isn't going to work well enough.

As Euclid indicates, an accident like this is a 'poster child' for actual derailment detectors, probably devices mounted on individual cars that can signal promptly when problems are detected.  To some extent such a device can also check or track bearing problems better than wayside IR detectors.

Personally I think any of the money Sarah wants to spend on ECP ought to be put dollar for dollar into this technology instead, until all the HHFT and PIH cars have been instrumented and the added programming (to PTC) for alerting and information exchange with the devices has been done.  I'd even go so far as to state implementation of these devices is a priority over the boondoggle that "PTC" has become at this point -- it certainly offers a much more positive and direct form of 'safety' provision.

[Euclid]

Maryville, TN Derailment:

Apparently an axle broke and derailed a tank car in a southbound train.  The broken axle was on the leading end of the car.  As the damaged truck settled down, I suspect that a rotating wheel may have then contacted the tank, burned through it, and ignited the escaping acrylonitrile.  As the train rounded a long curve and crossed Mt. Tabor Road, the derailed car ripped out the grade crossing for the road.  The train dragged the derailed car about 2000 feet further south to where it stopped short of Old Mt. Tabor Road.

I would like to know where the car first derailed.  Clearly, it was on the ground at Mt. Tabor Road.  The next grade crossing to the north (about 2700 feet north) is for Middlesettlements Road.  I don’t find any photos of crossing damage other than Mt. Tabor Road.  So perhaps the derailment happened south of Middlesettlements Road.  An eyewitness reported that the derailed car was on fire as it tore out the crossing at Mt. Tabor Road. 

Interestingly, the derailed car was misaligned with the train enough to foul the grade crossing signal at Mt. Tabor Road.  And yet, even with that much misalignment, the derailed car remained coupled and running with the train.

There are several photos looking down the track to the end of cars and clearly showing the fire.  These photos are looking north, depicting Interstate #321 overpass, and Old Mt. Tabor Road crossing barely visible beyond.  The train and fire are further beyond.  Most photos make the fire look rather small, but photos taken at night show very large fire.  It is amazing that it was prevented from spreading to the other tank cars.

I wonder if the train stopped due to the air hoses burning in two, or if the crew may have spotted the fire.

This was a derailment that luckily did not progress to a pileup, although it easily could have.  I am not sure where the other loads of acrylonitrile were, but the burning car had other tank cars trailing it.  If there were a pileup, those other tanks would have gone right into it. 

This would be a textbook case supporting the rationale for derailment sensors.        

[tdmidget]

AKA dragging equipment detectors. Quite possibly what alerted this crew. How many times are you going to reinvent the wheel?

 

[Euclid]

.

[schlimm]

Murray

 

 

schlimm

 

Murray

 

 

BOB WITHORN

Wouldn't it be nice to find the off switch!!!!!!!!!

 

 

 

 

Bucky is the "Eveready Bunny."  He keeps going and going and going and going and going......"

 

 

 

 

As do you.   You never miss a chance to belittle Euclid/Bucky.  Why can't you just ignore his remarks, the same as if there were a program on TV you dislike and change the channel?  Or do you throw a hard object at the screen?

 

 

 

 

There you go again Schlimm...always defending your friend and cohort Bucky.

You know Schlimm, the core group of us posters here had come to the conclusion that you were finally fed up with Bucky, and become a regular person.

Guess that isn't the case.

Perhaps with a little therapy you'll be able to come around.

 

I mostly dislike Euclid's posts and call him on it.  However, I do not engage in verbal harrassment of him, as you have done for years.  Your attempts at insults are regarded by me (and others) as mere nonsense as easy to ignore, much like the tantrums of a three-year-old child.  Your lack of understanding of the difference between free speech and ad hominem attacks is equally obvious.

[Euclid]

ECP braking does not offer much shorter stopping distance in an emergency, although it is somewhat shorter.  However, nobody here has said that shorter stopping distance is a significant benefit of ECP, so I fail to see why there is an argument about it.  What Wabtec would say about ECP is that it reduces UDEs, and reduces or eliminates slack action.  UDEs can be a great source of hard slack run-in, so there is double benefit in reducing UDEs, plus mitigating the slack run-in that they cause with conventional air brakes.  Wabtec will say that slack action can cause derailments. 

Going back to the emergency stopping distance issue:

Even though the stopping distance advantage of ECP is small, it may nevertheless result in significant benefit in terms of reducing the pileup effect of an oil train derailment.  Consider that this pileup occurs in a very short span of time.  So the time for mitigation is that short.  And while the ECP advantage is said to be only 7%, that head start may be more than enough time to mitigate the pileup effect.    

[Wizlish]

BOB WITHORN

So, as I try to pick through the clutter, ECP seems to improve braking in normal applications of slowing a train. In an emergency dumping the air is dumping the air, ECP or regular, it's dumping the air. So no real benefit in an emergency.

Buslist will have better technical information, but the 'best-case' difference is around 3%.  This may be due to slightly faster application time for the ECP valves (because they have better modulated control) or may reflect all the electric valves in the train dumping simultaneously vs. the sequential (at the speed of sound in compressed air) timing of the application in conventional air brakes as the emergency signal goes down the trainline.

Much the same effect of the '3% improvement' could be achieved by inserting magnet valves similar to the type in a FRED EOTD every so often in a consist, and opening these simultaneously just as the valve in the FRED is actuated.

The improvements for service braking are far more substantial, and the improvement provided by graduated release may be enormous.  The problem is that you cannot easily mix standards-compliant ECP and ordinary interchange cars in the same train -- the approach in the Lorenzo Coffin days, of just continuing the brakepipe through the cars without the 'right' brakes, would be inadvisable at best in today's litigious climate...

So the approach is to use ECP only in the dedicated services where its advantages 'pay'.  As noted, several Australian mining ventures voluntarily paid for ECP and swear by its benefits.  Were the benefits 'real' for unit HHFTs, it would certainly 'pay' to have them converted as Sarah wants... assuming that locomotives as well as full 'sets' of cars (likely in 20-car 'rakes') would only be used in oil service, and never need to be used for non-ECP service.  The chief issue the AAR sees is that the stated reasons to mandate ECP are not at all the reasons anyone would need ECP.  It is possible that evolving technology can provide a 'killer app' for this; my father clearly thinks it can.  But absent a really good silver bullet, the real-world benefits, from a very expensive investment, are really not going to be there...

[edblysard]

Yup, that pretty much sums it up.

[BOB WITHORN]

So, as I try to pick through the clutter, ECP seems to improve braking in normal applications of slowing a train. In an emergency dumping the air is dumping the air, ECP or regular, it's dumping the air. So no real benefit in an emergency.

[Wizlish]

Buslist or Euclid; I can't tell which any more the way it's formatted...

I would not accept any technical conclusions from studies on ECP at this point because the topic is so polarized due to its cost and a government mandate. There is too much riding on that dynamic to expect any technical information that is unbiased by the politics. The problem with the technical arguments is that nobody can confirm or deny them, and both sides have powerful reasons to make false claims.

But there is NO problem with the 'technical' conclusions regarding ECP so far -- as Euclid has already noted, the only meaningful contribution ECP could provide 'technically' depends on technology not yet developed (including presumably a good differential braking system and derailment sensors, etc.).  I thought everyone (except me) was more or less in agreement that only emergency-braking performance was of concern in stopping derailed or derailing HHFTs, and I have not seen data that indicate there is a statistically-meaningful difference in stopping time in emergency between the two systems, even before we start discussing the statistical likelihood of tank-car damage from an emergency stop vs. more controlled application or modulation. 

That is the context in which  AAR considers ECP economically unjustified given the supposed benefits for HHFTs (let alone as the camel's nose for mandated ECP in general railroad service, as I suspect Sarah Feinberg would mandate today if she could).  That is also the context in which the FRA espousement of ECP for HHFTs will have to be justified.

Interesting to compare this with the recent Supreme Court decision on mercury abatement and its cost vs. benefits...

[Anonymous]

schlimm

 

Murray

 

 

BOB WITHORN

Wouldn't it be nice to find the off switch!!!!!!!!!

 

 

 

 

Bucky is the "Eveready Bunny."  He keeps going and going and going and going and going......"

 

 

 

 

As do you.   You never miss a chance to belittle Euclid/Bucky.  Why can't you just ignore his remarks, the same as if there were a program on TV you dislike and change the channel?  Or do you throw a hard object at the screen?

 

There you go again Schlimm...always defending your friend and cohort Bucky.

You know Schlimm, the core group of us posters here had come to the conclusion that you were finally fed up with Bucky, and become a regular person.

Guess that isn't the case.

Perhaps with a little therapy you'll be able to come around.

[schlimm]

Murray

 

 

BOB WITHORN

Wouldn't it be nice to find the off switch!!!!!!!!!

 

 

 

 

Bucky is the "Eveready Bunny."  He keeps going and going and going and going and going......"

 

As do you.   You never miss a chance to belittle Euclid/Bucky.  Why can't you just ignore his remarks, the same as if there were a program on TV you dislike and change the channel?  Or do you throw a hard object at the screen?

[Anonymous]

BOB WITHORN

Wouldn't it be nice to find the off switch!!!!!!!!!

 

Bucky is the "Eveready Bunny."  He keeps going and going and going and going and going......"

[BOB WITHORN]

Wouldn't it be nice to find the off switch!!!!!!!!!

[Euclid]

Buslist,

I don’t doubt that experts you cite are experts.  I am just saying that there are two sides to this ECP debate, and each side will back up their points with experts.  Do you think the ECP experts are all going to admit they are wrong and the AAR experts are right?  Who is right when you have dueling experts?  Which ones am I supposed to agree with?    

I said that I disagree with the AAR position that ECP does not prevent derailments.  The wording of that AAR position is absolute, meaning can never happen.  I don’t think my inability to cite a derailment that would have been prevented by ECP proves anything, let alone that it cannot happen.  I agree that it might not happen often, but the AAR says it never can happen.  It is easy to disagree with that.    

To your point about the information from Sharma & Associates, you say they were attempting to predict the probable number of punctures in various tank car derailment scenarios, and you described their work as follows:

“They noted that there were so many variables and a wide range of potential values for each that an exact value was impossible to predict. The study used chaos theory (noting that derailments are chaotic events) to estimate a most likely result but noted for any given derailment an exact value was impossible to determine.”

I have said what I think about it.  What do you think about it?  Why did you post it?  What is their overall point of their work on this? 

What is your point in saying that “an exact value was impossible to determine”?  Is that a failure of the mission?  If that is what you are implying, what do you base it on?  They were not seeking an exact value.  You said they were attempting to predict the probable number.  What was the probable number that they predicted?  Can you please provide the full report of their work on this?  I would be very interested in reading it to learn the probable number of punctures in various tank car derailment scenarios.    

[Anonymous]

Norm48327

 

 

Buslist

Now I see why some get so frustrated with you. Experts be damned I know better.

 

 

Thank you.

 

Buslist....Truer words were never spoken!!!!!!!!

Now you know what the rest of us have had to endure with him.

[Norm48327]

Buslist

Now I see why some get so frustrated with you. Experts be damned I know better.

Thank you.

[tree68]

Buslist

Now I see why some get so frustrated with you. Experts be damned I know better.

[Buslist]

 

Now I see why some get so frustrated with you. Experts be damned I know better.

 

 

[quote user="Euclid"]

Buslist,

To be clear, I am not advocating ECP brakes alone as being a significant solution to the oil train derailment problem.  I favor them to be used in conjunction with further technology not developed yet.  I believe that would be a major contribution to the solution.  It has to do with getting the slack stretched at the start of a derailment, if it happens to be bunched at that time.

[\quote]

 

But the Chinese data suggests this may not be the case, and until that additional thing is developed you are advocating ECP as stand alone.

 

[quote]

But my point was in my response to Dave and his assertion that every little bit of added safety helps when it comes to delaying inevitable explosions by the use of thermal protection.  I disagree with that premise.  I am pointing out the inconsistency of him holding that “safer is better” position while denying the benefit of ECP by saying that it does not add enough safety to be worth it, as the AAR says. 

I would not accept any technical conclusions from studies on ECP at this point because the topic is so polarized due to its cost and a government mandate.  There is too much riding on that dynamic to expect any technical information that is unbiased by the politics. The problem with the technical arguments is that nobody can confirm or deny them, and both sides have powerful reasons to make false claims.

 

[\quote] 

 

Actual test data from disinterested parties is not to be trusted because it is biased in favor of who and for what reason? You may not understand that this not a political forum but a technical one. Papers are presented by industry experts and reviewed by a panel of experts before being accepted. In this case very few of the reviewers were Americans. The paper is then presented to a room full of industry experts (including some from FRA) and the results can be challenged publicly. These aren't some fluff papers. Perhaps next time around you can submit a paper with some data backing up your assertions. 

 

 

For instance, the AAR says flat out that ECP will not prevent derailments.  I think that is simply false.  

 

 

Are you serious? What derailments do we know of that would have been prevented by ECP? Certainly none of the CBR derailments as none appear to be train handling/dynamics related.

 

[quote]

I also believe that in many cases, ECP would mitigate a derailment once it begins, as has been roughly explained by Secretary of Transportation, Anthony Foxx right after the mandate. 

[\quote] 

 

And of course his vast experience trumps that of someone like Gary Wolf who as an independent derailment investigator has experience with thousands of derailments says it ain't so! I'll put my money with Gary.

 

I believe the industry draws a line in the sand saying that once a derailment begins; there is no way to mitigate it in process.  That view has been expressed here by various people.  It is a position that is immediately convincing to anyone who has seen the results of a derailment.  And it is a position that serves the industry by limiting their responsibility. You can hear it when the AAR says in effect:  “We think the best (and only viable) approach to the problem is preventing derailments.”

The issue I see with that is that there is no way to prevent all derailments, and with the consequence of an oil train derailment being so severe, it makes sense to look for a broader solution beyond simple prevention of derailments.  That is where ECP brakes come in.

 

 

With no concrete evidence of much benefit.

 

[quote]

And that is why the industry draws a line in the sand between prevention and mitigation. 

This line in the sand is quite evident in your reference to Sharma & Associates and the work they did on predicting the probable number of punctures in various tank car derailment scenarios. You say: “They noted that there were so many variables and a wide range of potential values for each that an exact value was impossible to predict. The study used chaos theory (noting that derailments are chaotic events) to estimate a most likely result but noted for any given derailment an exact value was impossible to determine.”

Clearly, this is to establish that the derailment process cannot be controlled or mitigated because it is chaotic.  I disagree with that conclusion.  You may not be able to make order out of chaos, but you can suppress chaos.  It is a smoke screen for them to conclude that it is impossible to predict the number of punctures in a tank car derailment.  Of course that is true, but it is completely beside the point. 

 

[\quote]

 

No it is exactly the point as you seem to be demanding precise numbers when terms of affection are all that are available. And why would Sharma want to put up smoke screen as this work was for the FRA.

 

[quote] 

Naturally the AAR is going to come up with every problem they can think of to make a case for no ECP mandate.

[\quote]  

 

 

So facts on why the industry considers the system unreliable is inappropriate in your view?

 

Not  only is there the burden of this tank car ECP mandate, but there is also the clear prospect of extending the mandate to all U.S. practice, now that the government has shown that it is willing to mandate ECP.  I think that is the 600-gorilla in the living room. 

 

 

 

As I believe has been explained to you several times before, the tank car situation was done under an emergency order which bypassed the usual protocol. To extend it would require an action of the senior RSAC committee, followed by a specialist committee to write the rule followed by an NPRM, and a review of the proposed rule by OMB for cost /benefit considerations, so FRA can't just decide to extend it as you imply.  

 

More likely the big question is how do I deal with smaller cuts of tank cars (non unit train).

Are they always placed right behind the locomotive, or do we equip non ECP equipped cars with ECP cable? How does the locomotive engineer deal with a train that has 2 different sets of brakes with differing operating characteristics?

[Euclid]

Buslist

Some interesting takeaways from sessions at the recent international heavy haul conference. In a conversation with Gary Wolf, probably the best independent derailment investigator in North America, he felt that the advantage of ECP brakes in emergency is so small that he felt the mandate was totally unjustified. This was further reinforced by a paper from China that, among other things compared emergency stopping distances of ECP and conventional braked trains for a variety of train sizes and speeds. On the graphs shown, the distance difference was almost imperceptible.              

 

In a bit of a surprise they showed a graph (that was not in the published paper) that seemed to show that longitudinal forces were higher in the ECP trains in emergency braking. Unfortunately the graphs shown had Chinese legends and the author was unable to understand questions in English, so that's a bit of a mystery.

 

The chairperson of the AAR air brake committee made a presentation on the state of the art in ECP. One of the issues related to unreliability of the system is cross talk between trains. Yes I know it's a wire line system, but when passing train A can hear messages leaking from train B and vice versa. If a locomotive detects a message from a car not in its consist it will initialize a penalty application. The spec is being revised so messages now include train number as well as car number. This is said to mitigate but not eliminate the problem as crossover messages will be ignored but will still consume coms capacity and may result in a time out penalty application. There have also been some issues with the connectors as they get older and wear, resulting in moisture penetration and ground fault failures. There is a search for a new connector.

It was noted that the committee feels that the tank car mandate will need to be an overlay resulting in dual systems on board the affected cars. The current cost estimate is in excess $6000 per car.

 

Sharma & Associates presented some work they did on predicting the probable number of punctures in various tank car derailment scenarios. They noted that there were so many variables and a wide range of potential values for each that an exact value was impossible to predict. The study used chaos theory ( noting that derailments are chaotic events) to estimate a most likely result but noted for any given derailment an exact value was impossible to determine.     

 

Buslist,

To be clear, I am not advocating ECP brakes alone as being a significant solution to the oil train derailment problem.  I favor them to be used in conjunction with further technology not developed yet.  I believe that would be a major contribution to the solution.  It has to do with getting the slack stretched at the start of a derailment, if it happens to be bunched at that time.

But my point was in my response to Dave and his assertion that every little bit of added safety helps when it comes to delaying inevitable explosions by the use of thermal protection.  I disagree with that premise.  I am pointing out the inconsistency of him holding that “safer is better” position while denying the benefit of ECP by saying that it does not add enough safety to be worth it, as the AAR says. 

I would not accept any technical conclusions from studies on ECP at this point because the topic is so polarized due to its cost and a government mandate.  There is too much riding on that dynamic to expect any technical information that is unbiased by the politics. The problem with the technical arguments is that nobody can confirm or deny them, and both sides have powerful reasons to make false claims.  For instance, the AAR says flat out that ECP will not prevent derailments.  I think that is simply false.  I also believe that in many cases, ECP would mitigate a derailment once it begins, as has been roughly explained by Secretary of Transportation, Anthony Foxx right after the mandate. 

I believe the industry draws a line in the sand saying that once a derailment begins; there is no way to mitigate it in process.  That view has been expressed here by various people.  It is a position that is immediately convincing to anyone who has seen the results of a derailment.  And it is a position that serves the industry by limiting their responsibility. You can hear it when the AAR says in effect:  “We think the best (and only viable) approach to the problem is preventing derailments.”

The issue I see with that is that there is no way to prevent all derailments, and with the consequence of an oil train derailment being so severe, it makes sense to look for a broader solution beyond simple prevention of derailments.  That is where ECP brakes come in.  And that is why the industry draws a line in the sand between prevention and mitigation. 

This line in the sand is quite evident in your reference to Sharma & Associates and the work they did on predicting the probable number of punctures in various tank car derailment scenarios. You say: “They noted that there were so many variables and a wide range of potential values for each that an exact value was impossible to predict. The study used chaos theory (noting that derailments are chaotic events) to estimate a most likely result but noted for any given derailment an exact value was impossible to determine.”

Clearly, this is to establish that the derailment process cannot be controlled or mitigated because it is chaotic.  I disagree with that conclusion.  You may not be able to make order out of chaos, but you can suppress chaos.  It is a smoke screen for them to conclude that it is impossible to predict the number of punctures in a tank car derailment.  Of course that is true, but it is completely beside the point. 

Naturally the AAR is going to come up with every problem they can think of to make a case for no ECP mandate.  Not only is there the burden of this tank car ECP mandate, but there is also the clear prospect of extending the mandate to all U.S. practice, now that the government has shown that it is willing to mandate ECP.  I think that is the 600-gorilla in the living room. 

[tree68]

Buslist

In a bit of a surprise they showed a graph (that was not in the published paper) that seemed to show that longitudinal forces were higher in the ECP trains in emergency braking.

This adds another factor that I, for one, hadn't been thinking about - what good does improved stopping distance do (assuming there is improved stopping distance, which also seems to be in question) if the track structure can't handle the increased forces involved?

We're theoretically trying to keep the train together, etc.  That's not going to happen if the track fails under the train due to the forces generated by the braking.

We would hope that the track is up to the task, but all it takes is one undetected weakness (or even a known weakness) in the structure and we've got a mess.

I think this comes under the heading "unintended consequences..."

[Buslist]

Some interesting takeaways from sessions at the recent international heavy haul conference. In a conversation with Gary Wolf, probably the best independent derailment investigator in North America, he felt that the advantage of ECP brakes in emergency is so small that he felt the mandate was totally unjustified. This was further reinforced by a paper from China that, among other things compared emergency stopping distances of ECP and conventional braked trains for a variety of train sizes and speeds. On the graphs shown, the distance difference was almost imperceptible.              

 

In a bit of a surprise they showed a graph (that was not in the published paper) that seemed to show that longitudinal forces were higher in the ECP trains in emergency braking. Unfortunately the graphs shown had Chinese legends and the author was unable to understand questions in English, so that's a bit of a mystery.

 

The chairperson of the AAR air brake committee made a presentation on the state of the art in ECP. One of the issues related to unreliability of the system is cross talk between trains. Yes I know it's a wire line system, but when passing train A can hear messages leaking from train B and vice versa. If a locomotive detects a message from a car not in its consist it will initialize a penalty application. The spec is being revised so messages now include train number as well as car number. This is said to mitigate but not eliminate the problem as crossover messages will be ignored but will still consume coms capacity and may result in a time out penalty application. There have also been some issues with the connectors as they get older and wear, resulting in moisture penetration and ground fault failures. There is a search for a new connector.

It was noted that the committee feels that the tank car mandate will need to be an overlay resulting in dual systems on board the affected cars. The current cost estimate is in excess $6000 per car.

 

Sharma & Associates presented some work they did on predicting the probable number of punctures in various tank car derailment scenarios. They noted that there were so many variables and a wide range of potential values for each that an exact value was impossible to predict. The study used chaos theory ( noting that derailments are chaotic events) to estimate a most likely result but noted for any given derailment an exact value was impossible to determine.     

[Euclid]

dehusman

 

Euclid

I quoted this comment above saying that the specification time cannot be used to predict the actual time.

 

 

 

I didn't read anybody doing that.

 

 

 

“The task force report noted that the AFFTAC results (time to tank failure) could not be used to directly predict tank car performance in actual fire conditions.”

 

 

If this is true, then there is no way to predict how much time first responders will actually have to act prior to an explosion. 

 

 

 

 

Correct.  The problem isn't the measure, the problem is you are (once again) expecting some exact value for something that is more or less a probability.

 

Your house probably has gypsum board walls.  They are probably required to be that material and a specific thickness by code.  Why?  For fire prevention.  They are a "thermal blanket" on the wall studs to retard the spread of fire.  They are probably 1/2 hour rated.  Does that mean in a fire that they will last exactly 30"?  Of course not.  They could last hours in some fires or they could last 15 minutes in others.  But in the standard tests, they can last 30 minutes.  A wall that only lasts 15 min in the standard tests doesn't provide as much protection.  A wall that lasts an hour provides more protection.  Fire escapes or walls designated as fire blocks have to have walls with a higher fire rating than regular interior walls.

 

With all the tank car design details it is the exact same thing.  They design the cars to meet certain standards.  Exceeding the standards is safer.  No one is going to give you specific numbers on "what that means".  It means its safer. 

 

 

So all this talk about giving first responders more time is moot because there is no way to determine that time or guarantee it.  How can first responders’ lives be protected by a safe time window, if the time is unknown? 

 

 

 

Reductio ad absurdum.  Obviously you are not a first responder.

 

Do we really have to explain to you why a car that takes 10 times longer to explode is safer?  Really?  You don't understand that? 

 

If I told you we were going to give you a choice of two fire suits, one that had a 100 minute fire rating and one that had a 1000 min fire rating, then have you put on the suit of your choice and stand in the middle of a burning room, are you telling me you would say that it was moot which fire suit you would pick because you don't know how hot the fire is and the exact number of minutes the suit would last in that exact fire, that you might pick the 100 min suit?

 

If you pick the 1000 min rated suit, then whatever process you used to reach that decision is the same process they used to say the 1000 min car is better.

 

 

Dave,

 

My comment about the time window not being guaranteed (which you say you did not read) is my interpretation of the quote in blue directly below your comment.  It is this:

 

“The task force report noted that the AFFTAC results (time to tank failure) could not be used to directly predict tank car performance in actual fire conditions.”

 

 

It is a direct quote from the NTSB Safety Recommendation dated 4/3/15: 

 

http://www.ntsb.gov/safety/safety-recs/recletters/R-15-014-017.pdf

 

 

 

Your analogy between delaying tank car explosions in a fire and delaying fire burning through drywall is weak.  The former includes the likelihood of explosions with no warning while the latter does not.  Managing your safe proximity to a fire is much easier than managing your safe proximity to a possible explosion of unknown magnitude.

 

Also, slowing the spread of fire by its natural heating and propagation (by fire resistant drywall) is much different than slowing the spread of fire that can propagate by an explosion.  In the first case, there is a finite amount of fuel being ignited on a gradual basis.  In the second case, a sudden explosion can add 30,000 gallons of superheated fuel to a fire that is almost burned out.  That then can lead to a second explosion, and a third, fourth, etc.

 

To use your analogy of standing in a room with a 100 hour fire suit versus a 1000 hours suit; it makes no difference if the fire lasts more than 1000 hours.  It is one thing to allow time for people to evacuate.  It is quite another thing to for first responders to go into the explosion range under the protection of a theoretical time window. 

 

You refer to the time window as a “probability” in defense of the admission that it is unpredictable in a practical sense.  Then you imply that I can’t see it the way first responders do.  I’ll bet you will find that most first responders are not going to want to play Russian roulette by working in the explosion range under the sole protection of a “probable” time window.  They get vaporized, and you assure them that the “probable” time window was better than nothing.       

 

I agree that every little bit helps.  Who can argue with that?  Well the AAR can certainly argue against that when it comes to ECP brakes.  You make a laughable reach when you poo poo ECP because it only works on ECP equipped trains.