Interesting…wasn’t that the same stretch of line that saw the terrible IC commuter wreck in the early Seventies where an older steel E.M.U train rear- ended a new aluminum double decker trainset with many injuries/fatalities? That accident also involved a misaligned switch, although IINM it was an automated switch and the accident resulted from a dispatching error…
From the GCOR.## 8.8 is intended to make sure that the switch, once thrown, is secure in the thrown position, and the trainman would have a harder time unlocking and throwing the switch under a moving car or train.## 8.4 is intended to force the trainman to the opposite side of the track from the switch, and with a train in-between him and the switch, prevent him from throwing the switch under the moving train…this is also part of the inspection process of passing train, the trainman is required by rule to give the passing train a roll by inspection.## ## While it may seem hard for non railroaders to grasp, it is quite easy, after being on duty for 6, 8 or 10 hours to get “lost” in what you are doing…our job requires us to think several steps ahead of the actual task at hand, and often, you think you have performed a task, only to find you did not, or, the opposite, thinking you didn’t do something you should have, only to find you did so a few steps ago…in a “panic” situation, it is easy to forget and it is easy to line a switch under a moving car, ask any yard man how many times he has done so, or started to do so, only to catch himself at the last minute.## ## In the Graniteville accident, the conductor was sure he had lined and locked the switch, in the investigation, both he and the engineer testified to that, and if you gave the conductor a polygraph asking about it, odds are he would pass.## Of course, he didn’t but in his mind was absolutely positive he did.## Same mind set caused the Shepherd Texas accident a few years ago…## The local switch crew backed their train into a siding next to the main street in Shepherd, released their time and track warr
Very well and completely said, Ed. [tup] Thank you for those practical, sobering, and realistic insights.
- Paul North.
This is the official ICC report on the accident.
http://ntl1.specialcollection.net/scripts/ws.dll?browse&rn=3649"
Pete
Hey guys, I’ve actually been to the New Mexico site and it is absolutely eerie. And it looks much as it did in 1956, telephone poles, wires, and all.
The siding is gone, and Automatic Block Signals repositioned accordingly, but everything else seems to be the same.
It should be noted that the head-on occurred in the neighborhood of 2 o’clock in the morning, as I recall, on a new moon situation of pitch blackness in the middle of nowhere.
Years ago I concluded that IF the engineer of the waiting passenger train in the siding HAD NOT tried to get the attention of the fireman at the UNLOCKED switch stand, everything would have been cool. So, maybe if someone does something stupid, unless his life is in immediate, direct danger, it might be better to let it pass until such a time as the matter can be discussed in a safer situation.
The paralleling north-south I-25 is about a quarter of a mile away, plus the old, two-lane roadway is still there about 100 to 200 feet away, both to the west of the track itself. The rail line is on a mild, downward, northward slope in a vast rolling hills area.
and for those of you who think that a electric switch lock will keep you from throwing a switch in front of a train. you are mistaken.
I concluded the same thing looking at it in retrospect, but there was no way that the engineer could have known that no lives were in danger, and that it would have been better to not try to communicate to the fireman.
It seemed that both men were drawn deeper into an error of misunderstanding by the actions of each other. And that error was begun by the fireman. If the fireman had not unlocked the treadle, or if the engineer had not reacted to it, or if the fireman had not reacted to the engineer’s reaction, the accident would not have occurred.
PTC, CTC, ACSES, intermittant cab signals, etc., will not prevent an event such as Robinson. Once the switch is put into hand, the switch machine is disconnected and the switch can be thrown. PTC (as well as almost all electronic train-control systems) will detect the open switch, but if the approaching train is closer to the switch than its braking distance to a full stop, the train will run through the open switch.
Some types of electric switch locks prevent the switch from being operated if there is an occupancy in the circuit, but those locks can be disconnected too. Railroads are eliminating electric switch locks as much as possible because they are high-maintenance. Some locations make it difficult to eliminate electric switch locks without expensive rearrangements of track geometry or a lot of money spent on signaling. A typical location where it’s hard to get rid of an electric lock is a mid-siding hand-throw crossover between a main track and a siding.
Any switch that can be power-operated makes an event such as
The FRA preliminary rules for PTC will still require electric switch locks for hand-throw switches in signaled territory if the maximum authorized speed is greater than 20 mph main track or 30 mph on a controlled siding, or a leaving signal in lieu of the electric switch lock. Otherwise switch locks and leaving signals will not be required. The electric switch lock is a device to prevent a hand-throw switch from being opened if there is an occupancy in the circuit, until time has run. It has nothing to do with a power switch.
As an aside to add to MC’s comments on GPS:
GPS is a method of approximating a train location to a physical track location. It’s an input device, just like a track circuit, a track transponder tag, or an axle odometer that counts off distance. Its good points is that it is cheap and available worldwide (most of the time). Its bad point is that its accuracy is fuzzy and that makes it vastly less valuable than most people imagine in a rail operating environment where inches matter and the operating technology is based on maintaining but not intruding into small distance tolerances between two or more moving trains.
For example, when a train stops at a signal indicating stop, the train is either NOT PAST the signal or PAST the signal. NOT PAST is safe, PAST is not safe. Having the GPS tell you “Dude, I think the train is sort of in the vicinity of the signal” is a singularly useless piece of information because you have absolutely no idea if the condition is safe or unsafe.
I don’t know of any technology with so much religious faith and magical
Absent some sort of accompanying act–such as drug or alcohol use, knowingly going to work without sleeping the night before, or something like that–I would be somewhere between shocked and offended were an incident like that to result in jail time today.
People get killed all of the time because of purely negligent acts–probably 100s per day on the highways alone. The same can be said for work sites that are similar to railroads. Although there are–just like there were in the 1950s–statutes that provide for “criminally negligent homicide” they are extremely disfavored and almost always involve some sort of reckless act that accompanies the negligence (like drinking, drug use, or sleep deprivation).
Moreover, how crazzy would it be to send someone to prison for a purely innocent misstake, when many first-time rappists, burglars, and the like often receive less than 2 or 3 years of actual exectued prison time for an act that intentionally inflicted or attempted to inflict a significant injury on a person?
Finally, with regard to how many lawyers would be involved “today” with regard to the civil liability, there is a very erroneous but very wide-spread perception in America that no one was ever sued prior to 1970. The statistics regarding the number of lawsuits per capita, etc. and the salaries that lawyers received in 1950 as compared to today might surprise you a great deal. Like medicine, the golden age to be a general prac
GCOR 8.7 Clear of Main track Switch also pertains. At a meeting point when the train is in the siding, crewmembers must stay 150 ft from the main track switch
It’s a bit disconcerting how many intelligent and some very intelligent people fail to grasp the accuracy requirements for train control. Contrast this with aviation where a positional accuracy of a hundred feet can be an immense help in preventing mid-air collisions [edit: note can be as opposed to is]. While surveyor grade differential GPS can determine position to within an inch or so, the processing to achieve that accuracy is not compatible with real-time use.
- Erik
A few months ago, RWM pointed us to an article (author’s name forgotten) which resulted in a mid-air collision over the rain forest in South America (Brazil perhaps). The tracking and navigational ability of the auto pilot systems was very accurate.
In that accident, as in many such events, there are a number of events or mistakes which cascade to the final catastrophy.
“Negetive” momentum (for lack of a better term) has an enormous ability to spiral out of control, when conditions are ripe.
The original post discusses the human aspect of this accident. How does one handle that type of situation?
The science of the mind is very fascinating and equally complex.
ed
You make a rule requiring people to stay away from switches under circumstances where impulse-throwing-of-a-switch would be catastrophic.
Pete,
Thanks for posting that link to the wreck report. In reading it, I am surprised at the points it raises that seem to conflict with and/or raise questions about the story as told in Trains Sept. 1998 issue.
For instance, in reading that story, I had concluded that the fireman thought that the engineer was telling him that the switch was mistakenly lined into the siding, and the engineer wanted him to line it for the mainline. However, the wreck report says that the fireman said that he thought the engineer wanted him to line it into the siding, so that is what he did. There is a tremendous difference.
If the fireman believed that the engineer was telling him to line the switch into the siding, I cannot understand how the fireman could have
Will or do the FRA’s proposed PTC rules require electric switch locks for hand-throw switches in what we today refer to as non-signalled or ‘dark’ territory ? Or will the PTC installation be considered as effectively converting such territory to signalled territory anyhow ?
In my previous post, I was not addressing the case of a power-operated switch that had been disconnected and put into hand-operated mode, nor other kinds of ‘work-arounds’ or bypasses of the set-up of the electric switch lock - just the normal capability of the equipment - which seems to me to be yet another special situation. People can do lots of odd things, I suppose, and the designers and systems can anticipate and guard against many or even most of them - but not absolutely all.
- Paul North.
The thing I found strange about the report is there is no mention of hours of duty time for any of the crew. Could it have been fatigue of the crew? Could it have been that the fireman was new to this particular line and unfamiliar with his surroundings? Most other ICC investigation reports have on duty times and experience on the division of the accident. Some reports even tell what the train crews were doing before their tour of duty. What strikes me as odd about the accident is the unlocking of the switch before the oncoming train. Did he do this to save a minute of time after the meet? Why touch the switch before the meet? The points were lined for the main when he walked up to the switch stand. He should have seen that. Is there a rule stating that switch locks should be unlocked before a meet? If there is. Why?
Pete
The wreck report states the rules, which say that a switch must not be unlocked before a meet. In the case of this Robinson wreck, there were two locks, one for the switch, and one for the treadle, which was a foot operated mechanical latch for the switch stand mechanism. To throw the switch, required unlocking both the treadle and the switch stand, and then stepping on the treadle, and then throwing the switch.
I assume that the fireman unlocked the treadle to save that little bit of time while he had time to kill ahead of the meet. O
I have a question:
Suppose you are an engineer running a passenger train at night down the mainline at 79 mph, and approaching a meet with an opposing passenger train, which is waiting in a siding with its headlight turned off. Now suppose that train in the siding turns on its headlight when you are say 1000 feet away. What, if anything, do you do?
Our mainline switch movements are reported on the air by the crew member throwing the switch, to be repeated by the engineer. If it’s an over-and-back movement, with the switch to be restored to its starting position and me in a position to watch it, I may not lock it, but I will leave the lock in the hasp, more to ensure that the handle won’t pop up and allow the switch to move under the locomotive/cars than anything else.
The 20’/across the tracks rule notwithstanding, we often move to where the engineer can see us, which helps ensure that another set of eyes is making sure we don’t second guess ourselves. Another good reason for using hand signals…
I never walk completely away from an unlocked switch. Don’t want my keys to wander.