Oil Train

[Wizlish]

dehusman

Wizlish

I am more concerned that turning down the threshold on defect detectors might lead to something other than the desired effect -- what are the actions that have to be taken when a detector 'flags' an axle?

 

On many detectors they are interconnected so the system detects when an axle is trending hotter, that information is alerted, the dispatcher contacts the trains and the train sets the car out.

By extension and from what you said, the same procedure applies to impact detection.  The immediate question is whether 'dialing down' the detector threshold as the FRA release indicated would result in a serious increase in the number of cars needing to be set out enroute.

I suspect it isn't practical to lower thresholds 'just' for the 4-5% of the traffic that represents trains carrying at least one car of crude oil, although I suppose if the decision to set out cars is at a dispatcher's discretion much of the same effect can be obtained.  Presumably the detectors provide actual scaled impact-force data, with the 'threshold' being just a setpoint that triggers an alarm, rather than just tripping if the threshold shock is exceeded.  Do dispatchers have on-demand access to those data?

Are there wayside detectors that analyze the shock waveform and characteristics generated by individual wheelsets or wheels, and discriminate between different types of potential damage -- for example by using broad-spectrum microphones bonded to the rail?? 

I concur that 'improvements' in wayside detection represent an advantage for all traffic, not just oil traffic, but part of what I am wondering is whether 'improvements' mandated to (ostensibly) make oil traffic "safer" have positive, or negative, or mixed implications for the cost-effective flow of other traffic.

[dehusman]

Wizlish

I am more concerned that turning down the threshold on defect detectors might lead to something other than the desired effect -- what are the actions that have to be taken when a detector 'flags' an axle?

On may detectors they are interconnected so the system detects when an axle is trending hotter, that information is alerted, the dispatcher contacts the trains and the train sets the car out.

I am now wondering whether what is desirable here is actually a coordination between detectors (and some standard for calibrating them) so that there is knowledge when a wheel on a particular axle is deteriorating progressively somehow.

20 year old technology.  Its already in place on most class 1's.  It applies to all trains.  Oil trains are only about 4-5% of trains and oil isn't that dangerous so railroads want solutions that address a broader range of trains.

[Wizlish]

Euclid

Wizlish

The FRA says this:

 

“These are important, common-sense steps that will protect railroad employees and residents of communities along rail lines.”

 

It is meaningless without clarifying how much protection there will be. 

 

How exactly will lowering the speed limit to 40 mph protect people?

 

 

It keeps the kinetic energy and perturbing forces on track and train down, while permitting a semi-reasonable track speed in the areas involved.  The lower the speed, the lower by the square of the speed the initial accident forces would be.  Did you miss the part about "in high-threat urban areas"?  This isn't a 40-mph restriction everywhere.  (I'm expecting the red herring about how many of the fireball incidents so far have happened in these 'designated urban areas'; but imho it's a preventative measure even when the political motives are not considered.)

 

I was referring more to the other actions taken by the FRA, particularly the notification to be given to first responders, and upon more careful re-reading I retract some of my agreement that the steps are 'important'.

 

I had thought initially that the purpose of the first-responder notification 'enhancements' was to put information in the hands of the responders before they arrived at an accident site, so they could know what to expect, bring appropriate PPE, etc.  Looking more carefully, there is little if anything that assures that.  On the other hand, it does seem to make sure that Government investigators will get what they want within 90 minutes or else.  I had thought the idea was to extract the 'relevant information" from the manifest data (or whatever) and ensure that this was available to first responders ASAP, ostensibly while maintaining security (including 'commercial security') over other manifest or consist information.  That to me would be a 'common-sense' action.  I would have gone further.  There should be a standard format for the information in this 'report'; in fact, there should be several, optimized for different transmission media.  Tthere should be conventions for communications systems to ensure that requesting the report via a particular medium (e.g., two-way radio) gets an appropriate version of the information returned to the requestor.  Etc.

 

Likewise I sympathize deeply with the perception of broken wheels as a concern, and think that addressing the issue is 'important' and 'common-sense'.  However, I'm not sure that calls for better-trained inspectors is much of an answer, other than a political one. (Unless this is a new profit opportunity in the making to train and develop certifications for wheel inspectors...)

I am more concerned that turning down the threshold on defect detectors might lead to something other than the desired effect -- what are the actions that have to be taken when a detector 'flags' an axle?   I am now wondering whether what is desirable here is actually a coordination between detectors (and some standard for calibrating them) so that there is knowledge when a wheel on a particular axle is deteriorating progressively somehow.  [Note: assume this is only needed for unit oil train moves initially, so no concern over switched-out cars, etc.]  Might be more useful if additional anomaly detectors become more closely spaced on routes with oil trains... and the detectors are sensitive to more than just impact magnitude... but those are questions for another topic and perhaps another thread.

 

[Euclid]

 

The FRA says this:

 

“These are important, common-sense steps that will protect railroad employees and residents of communities along rail lines.”

 

It is meaningless without clarifying how much protection there will be. 

 

How exactly will lowering the speed limit to 40 mph protect people?

[Wizlish]

wanswheel

DOT press release about oil trains

http://www.fra.dot.gov/eLib/details/L16335

 

In my opinion, 'these were important, common-sense steps' years ago, long before Bakken and Eagle Ford traffic.  The question now is what new steps should and can be taken to solve the underlying problems, which are far greater than broken wheels and first-responder information...

[wanswheel]

DOT press release about oil trains

http://www.fra.dot.gov/eLib/details/L16335

[Wizlish]

Euclid

His point is that there is no reason to try to control a derailment when it is possible to just prevent them.

I don't think that was his point at all.  What I understood him to be saying was that there is little reason to concentrate, or obsess, on developing complex and expensive systems that may or may not mitigate severe oil-train accidents when the money and effort involved would be better spent on prevention.  I took the point of his 'three-alarm fire' metaphor to be illustrating this.

By extension:

He brought up the analogy to air bags earlier.  In applying his point about preventing derailments instead of controlling them, this would be how it would fit the air bag analogy:

There is no point in putting air bags in vehicles when the problem can be solved by careful and lawful driving.  And yet we have air bags because, although careful and lawful driving would render air bags unnecessary, it will never be effective enough to solve the problem.

The analogy I think he is making is more that you don't put in 13 air bags, active rear-wheel steering, TPMS, radio cruise control, active bumpers, Chobham armor and springs all over the car when a much better approach will be a combination of proper training, prudent operation, proper road and vehicle maintenance, etc.

Note that this does NOT say that some level of improvement in oil-train design is overkill, just that there comes a point -- dhusman thinks it comes more quickly than I do -- at which a derailment of any particular magnitude on an oil consist will snowball into danger REGARDLESS of any cost-effective measures you might apply to the train, and perhaps regardless of any measures that don't themselves price oil-by-rail out of any market if implemented.

Incidentally also, I have not suggested that the only acceptable outcome is zero fireball wrecks.  I have used the phrase, “reducing the problem to an acceptable level.”  My references to solving the problem 100%, or addressing the remaining 20%, are only as reference points to explain the logic.  I think that the objective will be met by reducing oil train wrecks to a level that the public accepts, rather than reducing it to zero.  Of course zero would cost infinite money and be impossible to achieve.

The problem is that the public perception IS that fireball wrecks should be, and must be, reduced to zero.  That's not just an opinion of the moment driven by self-serving journalism or whatever.  Politicians are going to use this as a golden opportunity to cram in all sorts of bells and whistles of legislation and unfunded mandates -- take the history of PTC legislation after a couple of MUCH less terrifying wrecks as a case in point.  And there is little advantage, or opportunity, in reducing the incidence of derailments by any of the honest means that is done.  (And a good Democrat will add that railroads should already have had their six-standard-deviations of quality control for track integrity in place already...)

 To me there's no question that oil trains can be made safer if by some chance they derail.  There's also little question that no amount of armoring the cars or their connections will accomplish much of that safety, or (in my opinion) throwing their brake valves indiscriminately into emergency as promptly and powerfully as even a discriminating derailment detector can manage will.

Any system applied to a train will apply only to oil trains, at least to start.  Any argument about applying components like ECP to all interchange cars, or perhaps even to entire locomotive fleets, is really a red herring.  It's possible that if ECP is mandated for oil trains the cost of a proper system, or one that is nominally compatible with 'regular' one-pipe brake systems, will be reduced enough to make it attractive for other unit or semi-unit services, but it is far less essential there than it is for the oil service.  Likewise any 'gizmo' system for discriminating derailments (up to five-nines or whatever confidence is needed) is not needed beyond oil consists, no matter how useful it might be now or in future on PIH hazards, liquefied gas fuel, or other places.

Where I see the near-term interest is how the cost of applying mandated systems to oil trains gets passed along to customers vs. absorbed by the carriers ... and perhaps more importantly, whether the mandated requirements are a proper inducement to the shippers to perform proper degassing, etc., of their volatile crude before shipping.  THAT to me is really the most important part of making these oil trains 'safer', and as often happens in this kind of situation, arranging the economics to make the 'right thing' the least expensive option ought to produce the right effects.

As long as we can keep it from being a grandstanding unfunded-mandate sort of arrangement!

[Euclid]

Larry,

I understand and agree with your point about diminishing returns.  My point that I think you are responding to is my response to deusman’s point on the previous page.  His point is that there is no reason to try to control a derailment when it is possible to just prevent them.  I think that is flawed logic.  He brought up the analogy to air bags earlier.  In applying his point about preventing derailments instead of controlling them, this would be how it would fit the air bag analogy:

There is no point in putting air bags in vehicles when the problem can be solved by careful and lawful driving.  And yet we have air bags because, although careful and lawful driving would render air bags unnecessary, it will never be effective enough to solve the problem.

And I do not believe his point was that detection and inspection will solve the bulk of the problem.  His point was that because preventing derailments is a higher level of prevention than limiting damage after a derailment occurs, we should just do that higher level.  That I think is flawed logic.

Also, my 80%:20% ratio is just an arbitrary number for illustration.  I do not assume that improving wayside detection and track inspection necessarily will solve the problem 80%.  So I would not assume that a cost benefit analysis would necessarily favor that safety measure.

Incidentally, I agree that the industry has to look at return on investment.  But while the investment is quantifiable, I doubt that any two people would agree on what the return is.  The “return” is just a game of chance, and as we have learned, that game includes the chance of enormous loss in death and injury, it also includes the potential loss of the oil business.  How much do you invest to limit your risk?

Incidentally also, I have not suggested that the only acceptable outcome is zero fireball wrecks.  I have used the phrase, “reducing the problem to an acceptable level.”  My references to solving the problem 100%, or addressing the remaining 20%, are only as reference points to explain the logic.  I think that the objective will be met by reducing oil train wrecks to a level that the public accepts, rather than reducing it to zero.  Of course zero would cost infinite money and be impossible to achieve.      

 

 

[ruderunner]

Good points Larry, and I do understand the 80-20rule. That said no one yet knows with any certainty how much derailment sensors may cost, or differential braking or whatever else may develop.

So which likely costs more, outfitting the fleet of cars and locos with a couple thousand dollars in parts, or relaying all track in the continent with concrete ties and weldedrail?

This is why I'm looking at things that dont require a fundamental changes to cars,infrastructure,or locomotive's. Ecp is a major change to braking.

And again this isn't limited to oil trains. These could be tested on coal or wheat trains too.

[tree68]

Euclid

 So if the top level measure only provides 80% of the solution, I think other measure are justified if they address the remaining 20%.

To turn that a bit - the 80/20 'rule' also applies to expenditures.  When comparing the top level solution to the remainder, I'd opine that dealing with the top level solutions will only take up 20% of the total cost of reducing the problem to an acceptable minimum.  The remaining 20% of the solution will thus take up 80% of the costs.

Oh my gosh!  Tree's saying we should just give up!

Hardly.  

But in business, the law of diminishing returns does hold sway.  At some point, the cost of fixing a problem exceeds the cost of dealing with the problem when it does occur.  

"Five nines" reliability is a common goal in many pursuits.  But five nines isn't good enough if that .00001 happens in your front yard.  I get that, and I'm sure everyone else does, too.

So we look at relative return.  Taking care of the big stuff - track, defect detection, you-name-it, benefits the entire operation, not just trains of crude.  Conventional braking systems handle train control just fine for almost all trains running today.   

As Bucky notes, there are myriad potential causes of derailments, so when do we say "good enough?"   Obviously a zero derailment rate is desirable.  I'm sure the railroads would like that even more than the public.  But when does the the cost of removing all potential derailments exceed the return on investment?

ECP has been around for a number of years.  Unlike electronic devices, and PTC, the incidents so far haven't sparked a similar knee-jerk reaction from any regulators for the use of ECP.  That could well be because ECP likely wouldn't have had any effect on the outcomes.

[Euclid]

I agree that if you can prevent derailments in the first place, it is the best way to solve the problem.  There is no point in trying to control them if you can just prevent them.

But the problem is so multi-dimensional that nobody expects one single approach to solve it.  Wayside detectors and track inspection will help, but can never solve the problem completely.  I think it is also clear that making tank cars stronger is not going to solve the problem completely.

So with the uncertainty of getting tank cars able to be high-speed crash-proof, and the present limits in lineside detection, what is the solution?  I see a solution gap.  The solution needs to be multi-faceted in order to reduce the problem as much as possible. 

The solution is not to focus only on the top level cause, and say that if we prevent that 100%, we will end derailments 100%.  The reason that is fallacious is that it will be impossible to prevent the top level cause 100%.

So let’s say the maximum wayside detectors and track inspections solve 80% of the problem.  Do you just stop there because if that top level measure worked 100%, it would solve the problem 100%?

When I suggest a range of measures in relation to ECP brakes, I am certainly not saying that they should be implemented instead of wayside detection and track inspection.  Everything that can help needs to be done until the problem is perceived to be reduced to an acceptable level.  So if the top level measure only provides 80% of the solution, I think other measure are justified if they address the remaining 20%.   

[ruderunner]

Dave, I'm not saying it needs yo be limited to only oil trains forever, just that they would be a good test bed. Eventually such systems could be added to the entire fleet.

And what's wrong with attempting to do something after a wheel has left the rail? Are you suggesting that we do nothing at all to prevent a minor derailment from turning into a major one?

And no I dont have any idea how much air a generator would use, hence I'm keeping options open. If you have hard facts please share.

And re definition of derailment, isn't that pretty much what I said. I think we agree here.

Re stopped trains, OK so a speed sensor would be beneficial, not a challenge.

Also perhaps we aren't discussing empty oil train derailment since they generally dont catch media attention. No big fireball makes for not catchy headlines. That doesn't mean empty oil trains dont ever derail.

You candle vs fire argument is interesting. You do realize that a large fire can be started with just one unattended candle? I'd like to put out the candle before the oil catches fire.

[ruderunner]

Murphy Siding

 

 

ruderunner

Re read your own posts Zugman.  Square wheels is a definite way to cause derailments.  Not only the sliding action tearing up the rails but the ultimate in floatsptooting tearing them up.  Once the rails are damaged beyond use, derailments are sure to follow....

Doing something you know is wrong is akin to sabatoge.

Gee whiz guys, I had to explain that to you?

 

 

 

 

    If you could somehow interpret that post that way,  I have to seriously doubt the thought process behind anything you would post.  Floatsptooting indeed!

 

 

Sorry, flatspotting, just a typo.

Zug was being facietous. I know that, but I simply wanted to be clear about my statement that doinsomething, even if its wrong does not include doing things you already know is wrong. Not every suggestion is the best way to deal   with the situation but the only way to get to the best answer is to try a few things that may or may not work well, then refine the idea and try again.

[Murphy Siding]

ruderunner

Re read your own posts Zugman.  Square wheels is a definite way to cause derailments.  Not only the sliding action tearing up the rails but the ultimate in floatsptooting tearing them up.  Once the rails are damaged beyond use, derailments are sure to follow....

Doing something you know is wrong is akin to sabatoge.

Gee whiz guys, I had to explain that to you?

 

    If you could somehow interpret that post that way,  I have to seriously doubt the thought process behind anything you would post.  Floatsptooting indeed!

[dehusman]

ruderunner

That said:that's a MOW problem and I think thy are doing a good job as is.  Could it be better? sure but at what cost?

All the "solutions" on this thread address what to do AFTER the cars are derailed.  The welded rail examples eliminate causes so the derailment doesn't happen in the first place.  Since track problems are a major cause of derailments, eliminating track problems is a major accomplishment.  Installing all the stuff you guys want on one set of equipment benefits that one set of equipment.  Improving the track benefits every car of every type that passes over it.

Questions for all:  Since someone pointed out that the air generators for the derailment sensors may consume too much air, I've rethought using axle mounted generators.

Then again you guys are not operating with any data so you don't have any idea whether that is true or false or how much air one of those generators consume.

DH: I see we disagree on what constitutes a derailment.  But I'm unclear on your examples, especially of the wheels getting between the rails.  Wide guage I understand ( or rolled over rail etc) but you seem to imply that the wheels can move closer together on the axle.

A derailment is anytime the wheel leaves the rail.  Very simple concept.

A wheel can come loose and move inward.

How long does the average train car take to enter/pass a crossover/switch/transition?

Could be hours if its stopped.

DH: just because one truck has derailed doesn't automatically mean the follwong cars wil also derail.  If it due to poor trackwork, then yes the likelyhood is higher but it's not a given.  Review Trre68 story about the roadrailer that rerailed on its own.  It's not anisolated incident unusual yes but not a freak of nature.

100% of derailments start with only one truck derailing.  Only a very few end with just one truck derailing.  Yes having one truck derail is not unknown.  there was one las week.  But in that case, as in the Roadrailer incident and in the vast majority of all of those type of derailments, the derailed car is an empty (or a very light piece of equipment).  As i have pointed out several times now, NONE of the oil train derailments we are discussing involve empty cars, ALL of them involved loads, ALL of them involved multiple cars and trucks derailing.

And I suppose I'll chime in on my views of the inter train forces under braking.  Lets take the fact that we have discussed that car brakes are designed to maximize braking force for empties.  And by the same token that loaded cars are effectively underbraked because os that.  That means 2 empties are going to have similar deceleration rates under braking, and 2 loads will have similar decelerations rates under braking  but a load will have a much lower rate of decelartion than an empty.  And if a load is behind an emty, the load will run into the empty and try to push it.

How many oil train derailments have been caused by the slack running in?  ZERO.
How many derailments a year are caused by slack during heavy braking?  A tiny, tiny fraction.  Less than 1%.

You guys are trying to figure out how to put out a candle while there is a 4 alarm fire buring right behind you.

So lets build 3 trains, one a unit train of empties, onea unit of loads and one mied freight of empties and lods.  Which would be the most likely to derail in an emergency stp

 That's real easy.  The mixed freight. would be most likely.  If built properly according to the rules maybe one or two out of every thousand, couple thousand stops you would get a wheel on the ground.  Of course since you were in the process of stoping it would only be one or two cars so it wouldn't be a major derailment.

 

[zugmann]

ruderunner

Re read your own posts Zugman.  Square wheels is a definite way to cause derailments.  Not only the sliding action tearing up the rails but the ultimate in floatsptooting tearing them up.  Once the rails are damaged beyond use, derailments are sure to follow....

Doing something you know is wrong is akin to sabatoge.

Gee whiz guys, I had to explain that to you?

 

 

If you think that was a serious suggestion, then I am just going to step out of this conversation.  You are either trolling,  or I don't know what....

 So yeah, gee-whiz.

[ruderunner]

Re read your own posts Zugman.  Square wheels is a definite way to cause derailments.  Not only the sliding action tearing up the rails but the ultimate in floatsptooting tearing them up.  Once the rails are damaged beyond use, derailments are sure to follow....

Doing something you know is wrong is akin to sabatoge.

Gee whiz guys, I had to explain that to you?

[ruderunner]

Euclid

This is a system of oil train enhancement to increase safety.  It has four stages that are interrelated:

1)    ECP brakes with wire control.

2)    Enhanced load braking force combined with empty/loaded sensors.

3)    Derailment sensors.

4)    Differential braking.

 

These are the specific benefits:

Item #1 reduces derailments, mitigates pileups, mitigates tank car breaching, and is the platform for items #2-4.

Item #2 mitigates pileups and mitigates tank car breaching.

Item #3 mitigates pileups and mitigates tank car breaching, and is the platform for item #4.

Item #4 mitigates pileups and mitigates tank car breaching.

 

For item #4, I am not suggesting a major reduction in braking that will significantly increase the stopping distance.  But that notion seems to have taken hold in this discussion.  The concept only basically assures that there is less braking ahead of the derailment than behind it.  It is intended to address the derailment in the dragging car phase, before a pileup begins.  Not every first car to derail immediately does enough track damage to derail the following cars. 

But even though item #4 increases stopping distance somewhat, item #2 does the opposite.  So with this whole system, the stopping distance may be less than with the present air brakes and practice; and yet it will also baby the derailed car rather than jackknife it or stand it on end.    

 

OK I'm going to jump in again here.  I agree with Eiclids idea, especially the differential braking aspect.  But I temper that with the fact tha ECP is not widely accepted yet (in 20 years who knows?)

To catch up on post since I last had a chance to actually type...

Tree68: no personal attack felt, but sometimes we need to think outside the box(car)... Whats your opinion on how much truck deflections is acceptable?

Northwest: good point but how does that compare todoing nothing?  Is status quo acceptable?

Dhuesman and Zugman:  yes I understand that emergency is currently yhr fatest way to stop s trsin.  But what if we could make it stop faster? Load empty valves allow that. 

DH you examle of welded rail in long lengths is mostly correct in asmuch as it minimizes (not ELIMINATES) the risk of bad welds and weak joints.  Nor does it eliminate the risk of bad rail or poor installation.  A step forward but not infallible.  That said:that's a MOW problem and I think thy are doing a good job as is.  Could it be better? sure but at what cost?

Oltmand: I think you understand my point about load empty valves. But I have to aske, what is the current  "failsafe" of those valves?  Is it maximum braking?  is it empty limited?  And yes ECP could overcome the problems now associated with L/E vales but I think thats a long time coming.

Questions for all:  Since someone pointed out that the air generators for the derailment sensors may consume too much air, I've rethought using axle mounted generators. Or more preciecsly sideframe mounted.  Here's the question: is the three bolt bearing cap standard? Meaning is the patterns of those bolts the same regardless of truck/axle/bearing manufacturer?  My guess is yes it'sstandard and that would be a boon to my revised axle generator idea.

In regards to 23 degree curves and a derarilment at 10 mph, whats your opinion on how detremental that would be? Obviously not a giant flaming pileup, but certainly an inconvieniece.  But how dangerous?

DH: I see we disagree on what constitutes a derailment.  But I'm unclear on your examples, especially of the wheels getting between the rails.  Wide guage I understand ( or rolled over rail etc) but you seem to imply that the wheels can move closer together on the axle.  Isn't the axle machined with a step that would provide a positive stop for the wheels to prevent them from moving inboard?  By the same token, wouldn't the sideframes and bearings prevent the wheels from spreading? Not counting catastrophic failure of the axle or wheels I'm not seeing what you're getting at.

Provided the wheels are at the right spacing and the rails are too, how could the truck not deflect? 

And yes the trucks trucks will not be inline with the car when rounding a curve,but wuoldn't both trucks be deflected the same amount (not counting switches, crossovers and transitions) which is why I mentioned a timer.  The timier would allow for some distance of tavel into a crossover, transition etc where the system would not produce a derailment essage.  How long does the average train car take to enter/pass a crossover/switch/transition?

Remeber, I figured a "switching" mode which would mute the alarm during slow speed tight trackwork business.  This could be automated using an axle mounted genrator since that would provied a "built in" speed sensing circuit.  By built in I mean the control board coud sense ow sped operation and shut itself off but automatically rearm as speeds increase.

Euclid: I belive that most derailments star with one axle off the rails, provided that deraiment isn't caused by a rolled over rail or somting similar that throws the truck to the ground simultaneously. I agree that avoiding setting the brakes on a derailed truck would be beneficial, but that would require a redesign of the current brake system and likely multiple cylinders. Ideally you would want 1 cylinder per axle but even 1 cylinder per truck get cost prohibitive.

Wizlist: yes stopping the train safely versus fastest is the goal.

DH: just because one truck has derailed doesn't automatically mean the follwong cars wil also derail.  If it due to poor trackwork, then yes the likelyhood is higher but it's not a given.  Review Trre68 story about the roadrailer that rerailed on its own.  It's not anisolated incident unusual yes but not a freak of nature.

Euclid keeps going about ECP, and I agree with him that that is something that needs to happen.  Probaby not in the next 10 or 20 years, but maybe 30 years down the line.  The big advantage that ECP carries isn't just shorter stopping distances that are advertised now, but what could potentionally bencome of it.  Adding the basic electronic controls is the first step, then things like built in/automated load empty controls can be added, the Euclids differntia braking (which is similar to stretch brakingof passenger trains) could be added.  It's a whole new ball of wax with ots of possibilities.  Ju7st not very practical at this time due tocost, acceptance, and complexity.  I liken this to the advent of ABS in automobiles, the original systems were crude and not very effective.  But time and technology moved on and now the basic ABS system includes traction control (think wheelslip detection) and stability control (keeps you from spinning out/rolling the vehicle over).  These are analogous to derailment detection, improved emergency braking and even derailment avoidance.

And I suppose I'll chime in on my views of the inter train forces under braking.  Lets take the fact that we have discussed that car brakes are designed to maximize braking force for empties.  And by the same token that loaded cars are effectively underbraked because os that.  That means 2 empties are going to have similar deceleration rates under braking, and 2 loads will have similar decelerations rates under braking  but a load will have a much lower rate of decelartion than an empty.  And if a load is behind an emty, the load will run into the empty and try to push it.

So lets build 3 trains, one a unit train of empties, onea unit of loads and one mied freight of empties and lods.  Which would be the most likely to derail in an emergency stp?

[Murphy Siding]

zugmann

 

 

ruderunner

Zugman, do I really have to explaig that doing something you know is wrong versus something that might be wrong?

 

 

 

 

 

You could try telling me what you are talking about.  I have no clue.

 

 I'll second that.  I went back and tried to figure out what you were refering to and came up blank.  Can you explain please?

[zugmann]

ruderunner

Zugman, do I really have to explaig that doing something you know is wrong versus something that might be wrong?

 

 

You could try telling me what you are talking about.  I have no clue.

[ruderunner]

Zugman, do I really have to explaig that doing something you know is wrong versus something that might be wrong?

[Euclid]

This is a system of oil train enhancement to increase safety.  It has four stages that are interrelated:

1)    ECP brakes with wire control.

2)    Enhanced load braking force combined with empty/loaded sensors.

3)    Derailment sensors.

4)    Differential braking.

 

These are the specific benefits:

Item #1 reduces derailments, mitigates pileups, mitigates tank car breaching, and is the platform for items #2-4.

Item #2 mitigates pileups and mitigates tank car breaching.

Item #3 mitigates pileups and mitigates tank car breaching, and is the platform for item #4.

Item #4 mitigates pileups and mitigates tank car breaching.

 

For item #4, I am not suggesting a major reduction in braking that will significantly increase the stopping distance.  But that notion seems to have taken hold in this discussion.  The concept only basically assures that there is less braking ahead of the derailment than behind it.  It is intended to address the derailment in the dragging car phase, before a pileup begins.  Not every first car to derail immediately does enough track damage to derail the following cars. 

But even though item #4 increases stopping distance somewhat, item #2 does the opposite.  So with this whole system, the stopping distance may be less than with the present air brakes and practice; and yet it will also baby the derailed car rather than jackknife it or stand it on end.    

[Murphy Siding]

oltmannd

 

 

Euclid

Here is a mechanical derailment sensor that does appear to monitor the position of the truck rather than detecting vibrations.  Apparently the rectangular metal pan allows a feeler pin to move over its area, as the normal range of truck motion encompasses.  Somehow a derailment disrupts this physical relationship of the pan and feeler pin, which causes an emergency application of train brakes.    

 

I am not sure if that is also measuring movement of truck pivot.  But, in the derailment video segment, the truck pivot is involved and this device is reacting to it.  Actually, the way the device is positioned, it would have to have enough clearance to allow truck pivot whether it measured it or not.  The overall explanation is highly detailed, but not 100% clear in the sensing mechanics.    

 

https://www.youtube.com/watch?v=VEzsEQHkQ-Y

 

 

 

 

Take a guess what the number one cause of brake actuation would be on this system in NA?

 

 

  Hot air? 


(Sorry man- the made me do it.)

[oltmannd]

Euclid

Here is a mechanical derailment sensor that does appear to monitor the position of the truck rather than detecting vibrations.  Apparently the rectangular metal pan allows a feeler pin to move over its area, as the normal range of truck motion encompasses.  Somehow a derailment disrupts this physical relationship of the pan and feeler pin, which causes an emergency application of train brakes.    

 

I am not sure if that is also measuring movement of truck pivot.  But, in the derailment video segment, the truck pivot is involved and this device is reacting to it.  Actually, the way the device is positioned, it would have to have enough clearance to allow truck pivot whether it measured it or not.  The overall explanation is highly detailed, but not 100% clear in the sensing mechanics.    

 

https://www.youtube.com/watch?v=VEzsEQHkQ-Y

 

Take a guess what the number one cause of brake actuation would be on this system in NA?

 

[dehusman]

It's one thing to try to stop short of a collision or an obstruction on the track ahead of you, or to avoid a SPAD or ATC/PTC violation. There, scrubbing off as much kinetic energy as you can makes reasonable sense in most contexts. It is something quite different when you have the analogue to a tire or wheel coming off a vehicle you're driving at speed. Think about it...

I don't see as much of a difference as you do, I see them as much the same, the major difference is the SPAD/collision/obstruction is something in front of the train and the derailment is something within the length of the train. Part of the reason this whole discussion is so pointless is everybody's trying to figure out what to do with the car and the train after its already derailed, but very few people are looking at the derailment and its cause. You can't put the manure back in the donkey.

If you have a derailed car, you have had something bad happen to cause the car to come off the rails. That generally means the track is damaged or obstructed, or when the derailed car moves, it will damage or obstruct the track. Either way if the cars behind the first derailed car move, they will enter the damaged track area. That means they too will derail and also damage the track. The more cars you pour into the damaged area, the more the area will be damaged and the more chances there are the derailment will get progressively worse. This will happen whether the train slows quickly or slowly. Decelerating at a slower rate means each successive car hits the damaged area harder (more kinetic energy) and there are more cars that will hit the damaged area. Shoving more cars into a hole at a higher rate of speed will not improve the situation.

The only time stopping more slowly is beneficial is when the defect is not in the track itself and when the derailed car is not going to do a lot of track damage (such as an empty car with shelf couplers). In that subset of occurrences, stopping the train slowly is a better option. If the cars are loaded, the chances of increasing track damage are higher. None of the oil train derailments under discussion involved empty cars.

The trick to deciding when to put the train in emergency, of course, is detecting when the right conditions have been met and doing it soon enough that you can make a decision on whether to make a service reduction or an emergency reduction before the situation makes the decision for you.

Until that decision can be made in a timely manner, the system is set to stop everything as soon as it can, because in most cases that is a the better action.

[Euclid]

Wizlish,

 

That is right on.  I agree with your points 100%.  When the first car derails, it may drag for miles and never progress to a pileup; or it may pileup at the moment it derails.  Once a pileup begins, it may make sense to brake the train as hard as possible because the more kinetic energy that is dissipated braking, the less will be available to shove cars into the pileup.  The one exception would be if the cars ahead of the derailment happen to develop more braking force the cars behind it.    

 

But when the car is only derailed and dragging, I see a lot of risk that dynamiting the brakes in both directions from that derailed car will cause it to jackknife and begin a pileup.  It does seem like a boneheaded idea, as you say. Yet, we have looked at two different derailment sensors that do just that.  Of course, they offer simplicity, but it comes with a big risk of making a problem worse rather than better.

 

For this reason, I favor adding derailment sensors on top of ECP brakes.  Some seem to conclude that an emergency application is the best course of action.  It is a safe conclusion with conventional air brakes because it is the only course of action available. 

 

With ECP brakes, however, there are many options of brake application available besides just an emergency application.  Certainly ECP could withhold heavy application upon the sensing of a derailment, before the wire is broken.  But it can also withhold heavy application after the wire has broken at the onset of a pileup.  It is not locked into to an all or nothing response like conventional air brakes.

 

So ECP can react ideally to a derailment and dragging car, and also react ideally to the development of a pileup.  And also, because it reacts simultaneously to all cars in the train, there is less chance of producing in-train forces, which the more delicate braking is protecting against.     

[Wizlish]

dehusman

Any system that causes in train forces cause an empty car to pop off. All of these proposals that maximize the braking effort have the possibility of causing a derailment by inreasing the stress in the train.

But the purpose of the exercise is not to maximize the braking effort, it's to minimize the safe stopping distance.  With much more emphasis on 'safe' than on just 'minimized stopping distance'.

As with antilock brakes, the usual result is that the braking distance will be longer than if all the brakes on the train exerted their maximal achievable retarding force ... but the tradeoff is that train dynamics can be greatly improved or even optimized under given conditions.

I would argue that putting a train with derailed cars in emergency is already a bonehead idea.  The most that can be said for it is that you'd have plausible denial when the lawyers tried to say you were negligent for not stopping the train on a dime, and you have the flat wheels and jackknifes to prove your zeal to try achieving that. 

It's one thing to try to stop short of a collision or an obstruction on the track ahead of you, or to avoid a SPAD or ATC/PTC violation.  There, scrubbing off as much kinetic energy as you can makes reasonable sense in most contexts.  It is something quite different when you have the analogue to a tire or wheel coming off a vehicle you're driving at speed.  Think about it...

[tree68]

dehusman

Any system that causes in train forces cause an empty car to pop off.  All of these proposals that maximize the braking effort have the possibility of causing a derailment by inreasing the stress in the train.

Witness the story in Trains about the tank car that popped out of a consist on a grade on the SP Coast Line.  I think that was a recent issue - sometimes I dig out old issues are re-read them...

Were it not for the disconnected brake line, the report of a missing car from the receiving yard, and an observant track man, nobody would have missed it - it went far out of sight off the ROW.

And that was simple slack run-in.

[Euclid]

dehusman

 

Euclid

However, I have been informed that the only thing that can perturb a derailed, dragging truck to start a pileup is another derailed car. 

 

 

 

You have been grossly misinformed.  Cars pile up when the cars leave the track structure or otherwise get significantly out of line where the forces can start rotating the car or cars.

 

But I am not sure how to reconcile that because I have always heard that emergency applications can cause a derailment.  And I would think that a derailed, dragging truck would be particularly vulnerable to being disrupted into chaos by an emergency brake application.

That information came from you a few pages back.  So, if we set that aside, then I don’t need to reconcile it.  I understand your other points here about quicker stopping being a double edged sword.  ECP brakes would stop quicker, and when considered in a vacuum, that would increase potential in-train forces.   But ECP also fundamentally reduces in-train forces because of its simultaneous application.  So I suspect that overall, ECP reduces potential in-train forces.  Empty/loaded sensors would result in quicker stops, resulting in potentially greater in-train forces, but if combined with ECP, I would expect that to mitigate the higher in-train forces from quicker stopping resulting from empty/load sensors.

 

When will we learn the exact causes for the four oil train derailments in the U.S. following the collision in Casselton, ND?

[Norm48327]