I haven't seen anywhere that the number of handbrakes that had been set has been determined by the TSB.
In reality the "normal" operation here by MM&A was a setup for failure. Instead of tying down a train (of any size, type or lading) in the flat yard at Lac Megantic, a onetime CP division point, the train was stopped near the single crewmember's favorite motel on a grade only seven miles away. It would take an almost impossibly conscientious person to walk down the train, climb up on a car from the ballast, fully set the hand brake, and drop off and repeat a dozen or more times, and then walk the 750 feet or so back to the locomotive, climb aboard, release the air brakes, test the hold, and then set the air and repeat if more hand brakes were required, when the run was almost over, the taxi was waiting for him, nobody was watching and he had been working almost 12 hours. Why go to all this trouble when he "knows" that the air brakes are set and will hold the train anyway?
rcdrye In reality the "normal" operation here by MM&A was a setup for failure. Instead of tying down a train (of any size, type or lading) in the flat yard at Lac Megantic, a onetime CP division point, the train was stopped near the single crewmember's favorite motel on a grade only seven miles away. It would take an almost impossibly conscientious person to walk down the train, climb up on a car from the ballast, fully set the hand brake, and drop off and repeat a dozen or more times, and then walk the 750 feet or so back to the locomotive, climb aboard, release the air brakes, test the hold, and then set the air and repeat if more hand brakes were required, when the run was almost over, the taxi was waiting for him, nobody was watching and he had been working almost 12 hours. Why go to all this trouble when he "knows" that the air brakes are set and will hold the train anyway?
Most real RAILROADERS do. Yes, I must admit there are those who do take chances and short cuts, but most of us don't. Unfortunately, it's the ones that don't do their job (not saying this is the case for this incident.) and something happens and then the whole world paints every employee with the same brush.
Besides being conscientious, there is another reason we do our jobs. Self-preservation. Managers can and do go out and check trains for proper securement.
I take offence at this post, not that I think it should be removed or locked.
Jeff
jeffhergert Bucyrus jeffhergert Bucyrus Why does the Canadian TSB say that is unreliable? Maybe because it's the simplest explanation. Since the procedure was followed and it moved anyway, the test must be flawed or unreliable. Jeff, Thanks for that clarification about adding a safety factor. Regarding the statement by the TSB of Canada, when you say that “maybe because it’s the simplest explanation,” do you mean explanation of the Lac-Megantic runaway? Actually the TSB made their statement in 2009, so it is not a response to the Lac-Megantic wreck. In 2009, they said this: “Because it is impossible to verify hand-brake effectiveness by pulling or pushing cars on high grades, locomotive engineers cannot accurately know that management's expectations have been met every time cars are secured in accordance with CROR Rule 112.”I mean in response to the TSB's statement when ever it was made. They must figure if it (P/P test) was done and didn't work, it must therefor be unreliable.Jeff
Bucyrus jeffhergert Bucyrus Why does the Canadian TSB say that is unreliable? Maybe because it's the simplest explanation. Since the procedure was followed and it moved anyway, the test must be flawed or unreliable. Jeff, Thanks for that clarification about adding a safety factor. Regarding the statement by the TSB of Canada, when you say that “maybe because it’s the simplest explanation,” do you mean explanation of the Lac-Megantic runaway? Actually the TSB made their statement in 2009, so it is not a response to the Lac-Megantic wreck. In 2009, they said this: “Because it is impossible to verify hand-brake effectiveness by pulling or pushing cars on high grades, locomotive engineers cannot accurately know that management's expectations have been met every time cars are secured in accordance with CROR Rule 112.”
jeffhergert Bucyrus Why does the Canadian TSB say that is unreliable? Maybe because it's the simplest explanation. Since the procedure was followed and it moved anyway, the test must be flawed or unreliable.
Bucyrus Why does the Canadian TSB say that is unreliable?
Why does the Canadian TSB say that is unreliable?
Maybe because it's the simplest explanation. Since the procedure was followed and it moved anyway, the test must be flawed or unreliable.
Jeff,
Thanks for that clarification about adding a safety factor.
Regarding the statement by the TSB of Canada, when you say that “maybe because it’s the simplest explanation,” do you mean explanation of the Lac-Megantic runaway? Actually the TSB made their statement in 2009, so it is not a response to the Lac-Megantic wreck. In 2009, they said this:
“Because it is impossible to verify hand-brake effectiveness by pulling or pushing cars on high grades, locomotive engineers cannot accurately know that management's expectations have been met every time cars are secured in accordance with CROR Rule 112.”
They did make the statement about the unreliability of the push-pull test in a report on an earlier runaway, but they did not link the issue to that specific runaway. So it is unclear why they even mentioned it there in addition to the question of what they mean by saying that the test is unreliable.
In that runaway, the engineer started having trouble controlling the train on a down grade, so he made an emergency application. Once stopped, that called for setting handbrakes to hold the train during the recharge. The minimum number of handbrakes called for was 12. He set 35, but the train started to roll before he got back to the cab. He made it into the cab and put it into dynamic braking, but that did not hold it. The train finally stopped when the track leveled out. So the push-pull test never came into play during the incident.
The TSB comment is in this report of the runaway:
http://www.bst-tsb.gc.ca/eng/rapports-reports/rail/2011/r11q0056/r11q0056.asp
The comment is at the end of the section called, Train securement between Bybee and Tika.
rcdrye I haven't seen anywhere that the number of handbrakes that had been set has been determined by the TSB. In reality the "normal" operation here by MM&A was a setup for failure. Instead of tying down a train (of any size, type or lading) in the flat yard at Lac Megantic, a onetime CP division point, the train was stopped near the single crewmember's favorite motel on a grade only seven miles away. It would take an almost impossibly conscientious person to walk down the train, climb up on a car from the ballast, fully set the hand brake, and drop off and repeat a dozen or more times, and then walk the 750 feet or so back to the locomotive, climb aboard, release the air brakes, test the hold, and then set the air and repeat if more hand brakes were required, when the run was almost over, the taxi was waiting for him, nobody was watching and he had been working almost 12 hours. Why go to all this trouble when he "knows" that the air brakes are set and will hold the train anyway?
Why do it? It is the mans job! It is the job his employer expects him to do! It is the job he is being PAID to do! Railroads do not expect impossibly conscientious employees - they expect employees that comply with rules and special instructions when handling and/or securing trains. It is not rocket science, brain surgery or slave labor - it is railroading 101.
Why not park the train 'in town' where the terrain is level? I suspect, the carrier has received numerous complaints from local residents for the noise and or smell of having the engine running to keep air on the train and responded by parking 'outside of town'.
Never too old to have a happy childhood!
jeffhergert rcdrye In reality the "normal" operation here by MM&A was a setup for failure. Instead of tying down a train (of any size, type or lading) in the flat yard at Lac Megantic, a onetime CP division point, the train was stopped near the single crewmember's favorite motel on a grade only seven miles away. It would take an almost impossibly conscientious person to walk down the train, climb up on a car from the ballast, fully set the hand brake, and drop off and repeat a dozen or more times, and then walk the 750 feet or so back to the locomotive, climb aboard, release the air brakes, test the hold, and then set the air and repeat if more hand brakes were required, when the run was almost over, the taxi was waiting for him, nobody was watching and he had been working almost 12 hours. Why go to all this trouble when he "knows" that the air brakes are set and will hold the train anyway? Most real RAILROADERS do. Yes, I must admit there are those who do take chances and short cuts, but most of us don't. Unfortunately, it's the ones that don't do their job (not saying this is the case for this incident.) and something happens and then the whole world paints every employee with the same brush. Besides being conscientious, there is another reason we do our jobs. Self-preservation. Managers can and do go out and check trains for proper securement. I take offence at this post, not that I think it should be removed or locked. Jeff
23 17 46 11
Semper Vaporo Just asking a question here? If you were to set some brakes on a series of cars and then do the push/pull test, would re-checking the "tightness" of those that were set show that they are still set as they were before? As I remember of what I have seen of the braking system, there are a lot of chains and eyeloops involved in the linkages from the brake wheel to the brake hangars. Chains and eyeloops are notorious for not extended to the true possible length of the series of loops... rust and imperfections in the contact surfaces tend to seize before two segments reach the deepest interface. Thus, they can be brought to some tension level, but when jarred severely they can then slip past that imperfection and allow some small amount of slop in the tension. Does the test alter the results?
Just asking a question here?
If you were to set some brakes on a series of cars and then do the push/pull test, would re-checking the "tightness" of those that were set show that they are still set as they were before?
As I remember of what I have seen of the braking system, there are a lot of chains and eyeloops involved in the linkages from the brake wheel to the brake hangars. Chains and eyeloops are notorious for not extended to the true possible length of the series of loops... rust and imperfections in the contact surfaces tend to seize before two segments reach the deepest interface. Thus, they can be brought to some tension level, but when jarred severely they can then slip past that imperfection and allow some small amount of slop in the tension.
Does the test alter the results?
Yes. In fact, when you do a "golden shoe" test on a new or modified freight car design, you go around banging on the rigging to make sure you've handled all of the issues you mention.
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
Has it occurred to anyone here that one or more of the handbrakes could have been released by a vandal?
Also, with 11 brakes tightly/properly set, pushing the train in the direction of the downgrade would very likely cause it to move. I have personally seen yard crews shove cars deeper into a yard track without releasing the handbrakes on them, just because it would be inconvenient to walk the cut ands release them. So it IS possible to move a cut with the handbrakes set.
And, as to how tight it tight enough, I've always been told by T&E personnel that if you grab the chain and try to shake it, it is fully set if the chain cannot move sideways, not even a little bit, i.e., so tight that it feels like a solid piece of steel.
Just my two cents...
Yes, it is possible to move them with the locomotive, but keep in mind yards are flat, and with, say a pair of 1500hp switchers; it is still a struggle to move a cut with 4 handbrakes.
And the brakes in yards are set to absorb the shock of having cars kicked against them, so you don’t want them so tight they lock the wheel and slide…once you’ve broken the friction between the rail and the wheel, they will side till gravity and weight get them stopped….your goal is to make them behave like a shock absorber, not an immovable brick wall.
Securing out on the road is a little different, but the idea is the same, the brakes are used to hold the wheels in place, but so tight they lock the wheel, if it starts to slide, and it’s on a grade downhill….your goal is the same as with your automobiles ABS anti-lock brakes, locked wheels can’t stop the rail car or auto.
oltmannd Bucyrus “Because it is impossible to verify hand-brake effectiveness by pulling or pushing cars on high grades, locomotive engineers cannot accurately know that management's expectations have been met every time cars are secured in accordance with CROR Rule 112.” This is amazing. If the empirical test is unreliable, you don’t have a safe and reliable means to secure trains with handbrakes. "not being met EVERY time" <> reliable. It it possible to have a reliable system the does not work properly EVERY time. Block signals are a good example. You have to define the term "reliable", however.
Bucyrus “Because it is impossible to verify hand-brake effectiveness by pulling or pushing cars on high grades, locomotive engineers cannot accurately know that management's expectations have been met every time cars are secured in accordance with CROR Rule 112.” This is amazing. If the empirical test is unreliable, you don’t have a safe and reliable means to secure trains with handbrakes.
This is amazing. If the empirical test is unreliable, you don’t have a safe and reliable means to secure trains with handbrakes.
"not being met EVERY time" <> reliable. It it possible to have a reliable system the does not work properly EVERY time. Block signals are a good example. You have to define the term "reliable", however.
The term “unreliable” was my term, but it is a bit of a euphemism because I did not want to go too far out onto a limb. I agree that term, unreliable leaves open the question of degrees of reliability, and in the final analysis, nothing is 100% reliable.
However, the actual statement by the TSB was that it is “impossible to verify hand-brake effectiveness by pulling or pushing cars on high grades,”
Since the only point of the test is to verify, their statement that verification by the test is “impossible” has to mean that the test is so far from perfect that it is 100% worthless. So it fails to work properly every time.
So the TSB is stating that there is NO WAY for a train to be secured in Canada. How govermental of them.
BucyrusSince the only point of the test is to verify, their statement that verification by the test is “impossible” has to mean that the test is so far from perfect that it is 100% worthless.
But they didn't say that. You are guessing.
BaltACDSo the TSB is stating that there is NO WAY for a train to be secured in Canada. How govermental of them.
Perfect is the enemy of good, once again.
BaltACD So the TSB is stating that there is NO WAY for a train to be secured in Canada. How govermental of them.
The train was parked at Nantes because the "yard" in town is much too short for it to fit without blocking traffic. It also permitted starting the train on a down slope. There is a similar siding named "Vachon" a similar distance and elevation east of Megantic.
With approximately 60 cars held only by air brakes slowly leaking air, it is almost certain that one of them relaxed enough the create a slight "twitch" setting the entire train in motion. The train set still for 45-50 minutes: Newton's Law, " A body at rest. . ."
Why didn't the Engineer ask for advice or actual help? Hubris? Was he working after his hours had expired?
David K. Wheeler
oltmannd BucyrusSince the only point of the test is to verify, their statement that verification by the test is “impossible” has to mean that the test is so far from perfect that it is 100% worthless. But they didn't say that. You are guessing.
BucyrusGuessing about what? Are you referring to my conclusion that a verification test in which verification is impossible is 100% worthless?
Yes. You were putting words in their mouth. NOW you say it's "your" conclusion...
If a block signal is not 100% reliable, is it worthless? If so, why do we have them?
oltmannd BucyrusGuessing about what? Are you referring to my conclusion that a verification test in which verification is impossible is 100% worthless? Yes. You were putting words in their mouth. NOW you say it's "your" conclusion... If a block signal is not 100% reliable, is it worthless? If so, why do we have them?
I quoted their statement and I said what it “has to mean” due to their use of the term “impossible.” I did not attribute my conclusion of the test being worthless to the STB as though it was their statement. So I did not “put words in their mouth,” as you say.
The statement by the STB is not analogous to the fact that signals can fail. Their use of the term “impossible” would be analogous to saying that signals always fail. So, yes, if it is "impossible" for a block signal to work properly, then I would conclude that it is worthless.
Bucyrus oltmannd BucyrusGuessing about what? Are you referring to my conclusion that a verification test in which verification is impossible is 100% worthless? Yes. You were putting words in their mouth. NOW you say it's "your" conclusion... If a block signal is not 100% reliable, is it worthless? If so, why do we have them? I quoted their statement and I said what it “has to mean” due to their use of the term “impossible.” I did not attribute my conclusion of the test being worthless to the STB as though it was their statement. So I did not “put words in their mouth,” as you say. The statement by the STB is not analogous to the fact that signals can fail. Their use of the term “impossible” would be analogous to saying that signals always fail. So, yes, if it is "impossible" for a block signal to work properly, then I would conclude that it is worthless.
Really? There are more false clears in a year than trains that run away after being tied down.
oltmanndReally? There are more false clears in a year than trains that run away after being tied down.
Yes that is telling and I have no doubt that it is true. However, I am just presenting the position of the STB. They may very well be wrong. I would tend to conclude that they are wrong. Even four years after saying the verification from a push-pull test is impossible, they are still speaking about how the test is an essential part of rule 112.
Here is a new report from the Office of Rail Safety, apparently from the State of California. It focuses on the questions raised by the MM&A oil train runaway, including the questions that I have been asking here, including the number of handbrakes required to secure trains, and the method of testing that number. Obviously, the impact of the runaway on Canadian practice will spill over into U.S. practice.
http://www.cpuc.ca.gov/NR/rdonlyres/BD1A08EB-ECBF-4237-B3E2-17815E4CFA2E/0/PowerPointforthe072513CommissionMeeting.pdf
From the 2nd 'bullet point' on slide 18 of 22 of the Cal-PUC Power Point presentation linked by Bucyrus above (color emphasis in original; 22 pages/ slides, approx. 2.41 MB electronic file size in "PDF" format):
"It has been demonstrated that the push-pull test is not always a good indicator of whether an adequate number of hand brakes have been applied and not all handbrakes are effective even when properly applied."
- Paul North.
Its pretty simple, they push pull test is not the answer in this situation according to the rules anyway. The rules state you must set a sufficient amount of handbrakes to hold the train absent air brakes. Meaning, you have to set enough handbrakes to keep the train from rolling if the airbrakes fail after the train is left unattended. So you make an estimate of how many brakes are needed based on your experience on the territory. You set that many brakes and conduct a release test, meaning you release the airbrakes. If the train does not move after the slack has adusted then you are good, reapply the air brakes and leave. If it does not hold set a few more handbrakes and try again. Pretty simple really. This push pull test that seems to be the hot topic is for single car set out's/securement not when tying down an entire train or large cut of cars
Then there can also be site specific rules or policy's or procedures regarding the minimum number of handbrakes required at certain locations and there can also be further requirements and procedures for steep grade territories but these can vary from location to location and railroad to railroad etc, etc, etc. There can also be locations that rules say specific number of handbrakes is required, for example 1 brake for every 5 cars, and the rule can go on to say no release test is required on those tracks at those locations. The first paragraph is the basic standard and it can be more restrictive based on location and railroad. So, unless there is someone on this forum who works for the railroad in question, and on the territory in question there can really be no answer as to, "How many brakes would you have tied" except for "A sufficient amount to secure and hold the train."
Paul North posted a quote from the link I posted above in which the TSB of Canada outlines their concern over the procedure of securing trains on grades. In addition to the point that Paul posted about the TSB believing that the push-pull test is deficient, there are two more associated comments by the TSB of Canada that I have excerpted below.
Apparently they believe that the present method of securing trains is inadequate because the number of handbrakes needed is left to the discretion of the operating employees. Other information, I have seen indicates that the TSB wants to change the system so that the brakes are set according to a prescribed formula; a procedure that does away with the need for a push-pull test. The problem I see with that is that, even with a formula that finds the exact number of brakes needed, there is no way to know if those brakes are all performing to their full intended potential, and they often are not. The comments by the TSB also mention that problem:
Here is the excerpt, and the red highlight is from the linked document:
Transportation Safety Board of Canada Safety Advisories
*
CROR Rule 112 ensures that hand brakes are applied to prevent unwanted movement of the train while providing flexibility for a railway’s operating needs. However, CROR Rule 112 is not specific enough in that it does not indicate the number of hand brakes necessary to hold a given train tonnage on various grades and it continues to be left up to the operating employee to determine the number of handbrakes to apply.
It has been demonstrated that the push-pull test is not always a good indicator of whether an adequate number of hand brakes have been applied and not all handbrakes are effective even when properly applied.
Considering all these risks, Transport Canada may wish to review CROR Rule 112 and all related railway special instructions to ensure that equipment and trains left unattended are properly secured to prevent unintended movements.
The Rio Tinto Railroad is concerned about wasting time and effort by setting handbrakes unnecessarily. Their rules call for setting handbrakes on a train after its motive power has been cut off for over an hour with air brakes left holding the train. Their rules also call for setting a specific number of handbrakes to hold trains on grades. So they decided to see if they could fine-tune their handbrake requirements to save time and physical exertion on employees. They had two objectives:
1) To extend the one-hour limit to trains being held only with air brakes.
2) To reduce the number of handbrakes applied to hold trains on grades.
So, they conducted testing on their ore cars to learn exactly what the falloff rate of brake application force is over time. Ultimately, they were able to increase the one-hour interval to six hours.
They also conducted empirical testing on trains to find the exact number of handbrakes needed. This test was called the “Handbrake Holding Capability Trial.” This testing found that the number of handbrakes needed was considerably fewer than what the operating procedures called for.
The overall objective of this testing was to reduce the time spent in setting handbrakes, and to reduce the chance of personal injury involved with setting handbrakes.
Here are the details:
http://www.ceed.uwa.edu.au/ceed_seminar_proceedings?f=306870
It is not clear in the document, but I get the impression that Rio Tinto does not rely on an empirical push-pull testing, but rather, relies on a formula for setting a specific number of handbrakes that has been found adequate through extensive company research and testing. If so, that would differ from U.S. and Canadian practice. However, it might be the case that they are able to take the formulaic approach because they run captive rolling stock which they can maintain to top quality standards for brake systems.
It seems to me that handbrakes would need to be in near perfect condition in order to dispense with relying on train crew discretion through a push-pull test, and rely instead on the purely formulaic approach. It would not work to tell trainmen how many handbrakes to set if say half of the handbrakes were only capable of producing a portion of their full potential, and there was no way of knowing this by how they felt when being set.
Rio Tinto is not a common carrier and has several advantages as a result. It does not interchange cars with other roads, handles only one commodity and has a relatively homogeneous fleet of rolling stock.
There are many areas outside handbrake applications, where there are/have been struggles between those who favor a trial-and-error, individual judgement approach versus a formulaic or structured approach derived from empirical research. Adherents of both approaches can point out advantages and disadvantages of both.
C&NW, CA&E, MILW, CGW and IC fan
Question?
The requirement is to secure the train!
The specified formula is complied with to the letter.
After a period of time, the train rolls away.
Who is responsible for the roll away? Who will be charged for letting the train roll away?
BaltACD Question? The requirement is to secure the train! The specified formula is complied with to the letter. After a period of time, the train rolls away. Who is responsible for the roll away? Who will be charged for letting the train roll away?
That is a good question. The forumla has to have an adequatde safety factor to make it almost perfectly reliable. But anyting can fail if conditions are unsual enough. If it ever fails, the formula must be changed based on the failure experience. The ones who set the brakes to the formula would not be responsible.
But the biggest problem with the formulaic approach is the handbrakes themselves. Not only is there wide variation in their condition and effectiveness, but also, there is wide discretion in how tight to wind them. A better handbrake is needed.
Bucyrus BaltACD Question? The requirement is to secure the train! The specified formula is complied with to the letter. After a period of time, the train rolls away. Who is responsible for the roll away? Who will be charged for letting the train roll away? That is a good question. The forumla has to have an adequatde safety factor to make it almost perfectly reliable. But anyting can fail if conditions are unsual enough. If it ever fails, the formula must be changed based on the failure experience. The ones who set the brakes to the formula would not be responsible. But the biggest problem with the formulaic approach is the handbrakes themselves. Not only is there wide variation in their condition and effectiveness, but also, there is wide discretion in how tight to wind them. A better handbrake is needed.
The REQUIREMENT was to secure the train - the train rolled away and was thereby NOT SECURED. If the formula was followed or not is irrelevent, the REQUIREMENT was not met.
Our community is FREE to join. To participate you must either login or register for an account.