UP Trains Collide in Goodwell, OK

I've never seen a deadman pedal on a modern locomotive. Only first and second generation units. Anything newer would have an alerter, and even most first and second generation units have alerters.

Deadman pedals can be easily defeated without leaving evidence by simply setting something heavy on them. Alerters cannot be defeated by being kept continuously pushed in. You can defeat an alerter by cutting it out in the electrical cabinet. It is a serious, decertifiable offense to do so, and will show up on the event recorder.
UP units are equipped with event recorders that are automatically downloaded at trackside readers. If your alerter should fail enroute, that unit cannot be operated as a lead unit. Failure of an alerter must be reported to the dispatcher immediately.

The conductor, riding in the cab, plays the role of the old fireman in verbally calling out signals. Both the engineer and the conductor are required to do so, and to agree on what they are seeing.

I'm not real sure about what a "false clear" is, and how it would affect other signals. I do remember encountering a situation once on the BNSF with a signal I could not read. I was at Ceres Interlocking at the east end of the Sterling, Colorado yard. It is on the Brush, CO to Alliance, NE main. Ceres is where, eastbound, TWC ends and CTC starts. You have to get a signal to allow you to pull onto the controlled main.

At a certain time of year, the setting sun strikes the signals at such an angle that it is impossible to tell which one is lit. In fact, they ALL appear to be lit. Even leaving the locomotive cab didn't help.
The DS had given me a clear, but I could not tell that by looking at it. Finally, not wanting to guess at such a thing, I called him.
"I've given you a clear signal. Can't you see it?" He asked. "No. They all look lit up to me." I answered.
"Well, it is clear, believe me."
"OK." And away I went.

Now, I knew this was an optical illusion. But I also knew I could not tell what it was trying to tell me. So, in effect, it was a dark signal: one not to be passed. Calling the DS was all I could do.

Les

I am NOT saying that this is necessarily the case here, but it's entirely possible to keep resetting the alerter without actually being 'alert'. It's like when you keep hitting the snooze button on your alarm clock, then realize you've overslept... kind of a functional half sleep.
The gradual progression of intensity of the lights and sound, ease of resetting the system (horn, bell, throttle, brakes, in addition to the reset button) all contribute to an alerter not really ensuring alertness at all.

Sadly, just as it could have helped in the BNSF Red Oak, IA wreck, it seems that PTC could have prevented (or at least lessened the intensity of) this wreck too.

So sad to hear of the lives that were lost.

thirdtrick wrote:I'm not aware of any UP Cab Signal Territory outside of Chicago.

Back in the 1920s, the Interstate Commerce Commission ordered all of the Class 1 (large) railroads to install some kind of train control on at least 1 division. At that time, the Union Pacific didn't go to Chicago or Goodwell, Okla., but the Chicago, Rock Island & Pacific did go to both places. Some railroads equipped a remote and lightly traveled division, the Reading Company used Jenkintown - Bethlehem, not Philadelphia, to satisfy the order. While the Pennsylvania RR got an exemption to install Cab Signals on many divisions instead of train control on 1 division, some railroads just installed Automatic Train Stop without cab signals on 1 division. As traffic densities changed, some of these early installations were discontinued. That's why I wondered about the situation in Goodwell.

Desertdweller wrote:JayBee,

I'm having trouble reconstructing this in my mind. You seem to have the most information about what happened.

The Westbound Autorack was supposed to take siding at ESS Goodwell. At the Distant Signal they were given an Advance Approach and slowed to enter the siding. So far, so good. The Eastbound Container Train did not stop on the main at ESS Goodwell. That alone might not mean anything, depending on how close he was to ESS Goodwell. If he was within four signal blocks of ESS Goodwell, he should not have had a proceed (green) signal until the Westbound was in the siding at ESS Goodwell and the switch restored for through movement on the main. Of course, if that had happened, there would have been no collision.

You then state the collision happened on the main between Goodwell and Optima Siding. This would seem to me to be a head-on collision. Thus eliminating the circumstance that the Eastbound ran into the Westbound's train before he had fully cleared the main at ESS Goodwell. So, you say that the Westbound did not enter the siding at Goodwell? Or do you mean that the Westbound cleared the main at ESS Goodwell and re-entered the main at WSS Goodwell? If that is what happened, a properly functioning CTC system would not allow that.

Then you said that the Westbound Train had a meet with another Eastbound at Optima Siding. Is Optima Siding East or West of Goodwell Siding? Are you saying that the Westbound had gotten in the clear at Optima Siding, had a successful meet with an Eastbound, then pulled out to get smacked by a second Eastbound? How would he have gotten a clear signal coming out of WSS Optima with another train running into his face? More importantly, with CTC-controlled sidings, how could he have even been lined back onto the main in the face of an oncoming train?

Les

From what I read of that post, the second Eastbound train wasn't supposed to be past ESS Goodwell. Assuming no major malfunctions of the signal system this implies that this train in particular for whatever reasons had to have passed a red signal. The Westbound train contrary to news reports apparently never had a chance to take the siding for the meet, as the Eastbound train they were supposed to be meeting was already past the siding in question. A properly functioning CTC system much like any traffic control devices only works if all parties obey them, apparently one party didn't obey them.

My speculation based upon the time of day that the incident occurred is that the Eastbound train most likely had sunlight shining into the cab. This might affect the awareness to train control devices such as signals, which might explain how they missed signals requiring them to slow down and stop so that the Westbound train could take the siding. The fact that the Westbound train had an Advanced Approach implies right away the the siding at ESS Goodwell wasn't lined for the main ( it most likely would have been red if it were lined for the main track ).

If the CTC was working properly, the Westbound should have gotten a red signal at ESS Goodwell regardless of how the switch was lined. These signals consist of two sets of signal heads: the upper set displays indication for the main, the lower set indicates permission to enter the siding. Assuming the siding itself did not have occupancy signaled the train crew should have seen a red indication (for movement on the main) over a white (permission to enter the siding, but requiring the train crew to determine if the siding was clear). This would be equivalent to pulling into a track under the Yard Limit Rule.

The signal to enter the siding (Lunar White indication) would not be given until the mainline switch was lined and timed out. Only then would the light be illuminated.

Until the switch at ESS was lined and timed out for movement on the siding, the indication would be red with a dark signal for the siding. The red signal would be caused by the oncoming Eastbound.

Usually, in CTC, the signals will be dark until a train approaches from either direction. Signals are typically installed two miles apart in open country. The aspects will be lighted when a train approaches three or four blocks away. This allows an Advance Approach indication to be shown to indicate a red block two blocks away.

There is also a rule that requires a dark signal to be acted upon as if it is the most restrictive aspect that signal can give. Generally, that aspect would be a red signal. This rule prevents trains from blowing past malfunctioning signals.

Yes, you can thump an alerter button like the snooze button on an alarm clock. But trackside signals are required to be called out and acknowledged between the engineer and the other crewmember in the cab. This is why two-person crews are required on Class-One mainlines. If the CTC system were working properly, several
rules would have to be ignored to cause a wreck like this. This is why I suspect a failure of this system.

Les

Possibly this YouTube video will give some feeling as to "the lay of the land".

It does establish one fact; namely that the incident occurred at 1008A. This would suggest the sun was high enough so as not to interfere with the "C&E" from observing and reporting to one another on the signals'aspects.

While I do not condone that the driver was also videotaping, there nevertheless is a worthwhile information presented through this video:

http://www.youtube.com/watch?v=4t67iF9F ... re=related

this is one of the saddest stories i've ever heard in railroading.

Desertdweller wrote:If the CTC was working properly, the Westbound should have gotten a red signal at ESS Goodwell regardless of how the switch was lined.

From the information above, the westbound never made it to Goodwell. The crash occurred east of Goodwell.

Desertdweller wrote:If the CTC system were working properly, several
rules would have to be ignored to cause a wreck like this. This is why I suspect a failure of this system.

But those "several rules" would all have to be violated by the same crew.

From what's been posted above and ignoring the mass media stories which make little sense, this sounds to me to be another Chatsworth. The eastbound went by the stop signal and ran through the switch at the east end of Goodwell. The westbound was where it was supposed to me. One set of failures by one set of people (the crew of the eastbound). There might be something more involved but when there's a straight-forward explanation, that's usually it.

Some people who are thinking multiple failures (including some of these alleged experts the media is quoting) are assuming things work differently than they do. With CTC, the signals are the authority to occupy the track. There is no need for trains to be talking to each other (and gets in the way of required communication) and no need for the dispatcher to tell a train verbally what the signals will tell him (and also gets in the way of required communication).

The mass media story says a train failed to take the siding when it appears the reality is it failed to do so because it never made it that far. The real question is not why a train didn't enter the siding but why the EB and WB were nose-to-nose on the same stretch of single-track.

Desertdweller, just a nit but I don't know why you think time has to run normally to line a train into a siding. The only time time normally need to be run is when a proceed indication has been put up on a signal and then needs to be taken away. This is to insure safety when a train has acted on that signal or one that acts as an approach signal to it (e.g. train is lined up to hold the main at a passing siding, dispatcher changes his mind, takes down the signal. Before he can throw the switch, time must run out in case the approaching train has already seen and passed the approach signal indicating Clear rather than Approach). If a signal has never been put up, then the plant is unlocked for whatever the dispatcher wants to do. Under normal operations, the dispatcher does not put up a signal until he has a train to line up (in other words, signals at control points "rest" at Stop rather than a proceed indication).

Also, approach lighting of signals is independent of CTC. You can have approach lighting on ABS and always-on signals with CTC. Where I live now, the nearest railroad is always-on CTC while where I lived before that was approach-lit ABS.

NTSB has a few photographs up...


http://www.ntsb.gov/investigations/2012 ... ll_ok.html

Including one of the ESS at Goodwell, showing clearly to my untrained eye that it has been run through.

N_DL640A wrote:I am NOT saying that this is necessarily the case here, but it's entirely possible to keep resetting the alerter without actually being 'alert'. It's like when you keep hitting the snooze button on your alarm clock, then realize you've overslept... kind of a functional half sleep.
The gradual progression of intensity of the lights and sound, ease of resetting the system (horn, bell, throttle, brakes, in addition to the reset button) all contribute to an alerter not really ensuring alertness at all. .

Do we know for sure there was an alerter system here? IF so, how does it work - what/who actually transmits the signals to the alerter system. Is it an integral part of the CTC system, and whatever is telling which signals to light up is sending something to the locomotive? How is it actually transmitted -- some kind of code transmitted electrically through the tracks?
JS

pumpers wrote:

N_DL640A wrote:I am NOT saying that this is necessarily the case here, but it's entirely possible to keep resetting the alerter without actually being 'alert'. It's like when you keep hitting the snooze button on your alarm clock, then realize you've overslept... kind of a functional half sleep.
The gradual progression of intensity of the lights and sound, ease of resetting the system (horn, bell, throttle, brakes, in addition to the reset button) all contribute to an alerter not really ensuring alertness at all. .

Do we know for sure there was an alerter system here? IF so, how does it work - what/who actually transmits the signals to the alerter system. Is it an integral part of the CTC system, and whatever is telling which signals to light up is sending something to the locomotive? How is it actually transmitted -- some kind of code transmitted electrically through the tracks?
JS

The alerter is basically a timer. At a set time interval the device will go off (buzzer, blinking lights, etc). There are numerous inputs that satisfy it like brake cylinder pressure, throttle movement, blowing the horn.....I just found out this morning out of the blue on my own loco that keying the radio mic satisfies it. If no input is seen, then it initiates a penalty application of the brakes.

An "alerter" usually is a self-contained system in the locomotive that requires the the engineer to do something periodically to verify that there is an "alert" person at the controls. I've seen different forms: a switch to be pressed periodically, touching of the controls, and I'm sure there are others. It is not tied into the signal system or anything outside the locomotive.
Systems that have knowledge of the signal state are known by names such as cab signals, ATS (automatic train stop), and variations of that. Nothing posted has suggested that any system of that nature was used here. While there are exceptions, systems of that sort are mostly found in heavy passenger corridors and they are still the exception rather than the rule in the U.S.
As someone posted above, it is entirely possible for someone to be keeping the alerter happy while being oblivious to more important stuff like signals since keeping the alerter happy is something you do by rote while getting a signal you weren't expecting requires a conscious realization that the plan is changing ("hmmm, Approach. Guess I better start stopping this train").
(note: somewhat duplicitous of a post made while I was writing but enough new information to still be worthwhile).

The Alerter works off a variable timer, at very low speeds the time between Alerter activations is fairly long, at speeds in the 60 mph range it is about every 15 seconds. As mentioned, while the Alerter is an improvement over the Deadman's Pedal, it doesn't protect against instinctive reactions. The NTSB website has a photograph showing the ESS of Goodwell Siding with the Investigators measuring the gaps between the Switch Rails and the Stock Rails of the Switch. That fact that there is significant gaps on both sides is consistent with the switch being run through by the Eastbound Container train.

The only real issues to be discovered if possible;

Was there a Brake problem with the Eastbound
Or was the crew of the Eastbound train distracted, or Incapacitated, or Asleep.


The Data recorder from the DPU on the Eastbound should confirm or eliminate the first possibility.
They may never be certain about the answer to the second unless there is a record of a cellphone call.

That is not to say that there might not be other contributing factors.

While this "ditty" best described my eleven year railroad (held Non-Agreement positions within Accounting, Labor Relations, and Operations with a Class I) career, I am so pleased that this discussion has moved forth in such a mature and respectful manner, and that those of us not involved with the operation of trains have "stood aside" and let those who do or have done this stuff for a living carry the discussion.

While of course the outcome of this unfortunate incident with an even more unfortunate ending must await the final report from the NTSB, I am concerned about Mr. Stone's observation that "this sounds to me to be another Chatsworth". While of course, the foundation of this comment must await the noted outcome, all too many regulators and legislators are not known for their patience - especially when there are headlines to feed upon. We all know that (whatever meets the definition of) Positive Train Control has been mandated under RSIA '08 (legislation, lest we forget, signed by a Lame Duck president in response to Chatsworth and who I doubt if he even knew what he was signing let alone the consequences of such) PTC is mandated to be installed on any line, passenger or freight, that handles any appreciable volume of traffic. The implementation date is supposed to be 2015, but as often the case with mandates, there is "slippage'. While PTC on lines handling any volume of passenger traffic is inevitable, if not already there, there is the public trough to address that.

But the only way the industry will recover the costs of PTC on freight lines (such as through Goodwell) is through increases in rates that will make rail shipping less competitive in an environment where the competitive pressure will be on. This means in a post-PANAMAX world, ocean shippers will be looking for a maritime port as close to the final destination of their high value ex-im traffic, thereby depriving the industry of the favorable line hauls now enjoyed when traffic is routed through West Coast ports. Even "safe' traffic such as coal is under "pressure' as the environmental issues limit domestic use and export traffic could be affected if with higher rates export coal is no longer competitive when landed in Asia. Even crude oil being extracted from domestic fields where presently there are no pipelines will simply compel the producers to build such in the face of higher rail rates. There goes a source of traffic that would hopefully offset potential losses from export import coal and manufactured goods.

I'm as sorry as the next guy this tragic incident occurred and I too extend condolences to all concerned. But please, let this not be a "rush to judgement" with regards to implementation of PTC.

disclaimer: author holds long positions CSX KSU UNP

After scanning through the previous posts, I'd like to add some facts that I have not yet seen added to the discussion, as well as my two cents' worth:

- I don't work this particular territory, but from looking at the Google Earth satellite views, it appears that Goodwell is on tangent (straight) track, therefore the headlights on opposing trains would have been visible to each other for many miles. In theory, both crews could have seen each other, and would have known of an impending meet. In past years, I have worked in territory that was straight and flat, and I remember that I could see the headlight from an opposing train for many miles.

- The Z train crew would have been called out of their home terminal of Dalhart, which is only about an hour west of the wreck site. Even if the they WERE asleep or incapacitated, the lead unit did have an alertor, which would have applied the brakes and stopped the train if not acknowledged within about 20 seconds or so. Now with this in mind, I have developed such an automatic reflex, that I think I could hit and alertor button while half asleep.

- The morning sun COULD have been a factor. Even at 10 am, it would not have been directly overhead, so the crew may have not seen the wayside signals clearly. The current timetable shows it as CTC territory. If the signal system is set up as it is on other 70 MPH territory on this railroad, the "Z" train, if lined down the main, would have had a flashing yellow at the distant signal, a hard yellow at the WSS Goodwell, and a red at the ESS Goodwell. The railroad has a current rule in effect which requires crews to call out on the radio the aspect of approach and red signals. If I had been on the westbound auto train, seeing the approaching Z train's headlight, but not hearing the crew call the signals, I would have been on the radio with some kind of innocent talk directed to that train, trying to make sure they were awake! But that's just me - I'll tell you an unrelated story: 25 years ago, I was operating a train in ABS-TWC territory, and holding a no-meet Track Warrant all the way to my destination. At about 9:30 in the morning, while running on green signals at 50 mph, I heard a hotbox detector just 3 miles from my location fire off. Grabbing the radio, I called out for "any maintenance of way on the tracks located at the HBD at milepost xxx.x". To my shock, I got a reply from a Signal Maintainer, who told me he would be off the rail in about a minute! I plugged my train and stopped about a half-mile from his truck. After I cleaned out my shorts and double-checked my Track Warrant, I got off the engine and had a Come-to-Jesus meeting with that new Maintainer. He had a Track Warrant that was only valid "after arrival" of my train, but had mistakenly thought he could occupy the track until I got there.

- It is unfathomable to me to operate a train at 68 mph while staring into a headlight! But then again, in light of some of the more recent accidents, I have asked myself time and time again, "what were they thinking?" Even if the crew mistook a dimmed headlight for a yellow approach signal, this still reduces their authorized speed to just 30 mph.


I HAVE seen some posts that imply the CTC system actually "lines" or gives signals to trains, and that the Dispatcher can tell what color signal you have in the field. It does not, and he cannot. CTC is usually laid over an existing ABS block system. The only thing the TD can control are the "Control Points", and even at that, he can only line a power switch (assuming the switch isn't "in time") and allow or remove a more permissive signal indication at that CP. If a train, rail car, or track condition were present just east of Goodwell, the TD could attempt to line the eastbound train all the way through Goodwell, but the ABS signal system would NOT allow the signals to set up and display a permissive indication simply because the track circuit would be shunted. For these reasons, unless the signal system was faulty, I am confident that the eastbound train would have had a flashing yellow into a yellow into a red at Goodwell. The westbound auto train would have had either a flashing yellow (depending on the speed of the turnout at Goodwell) or a double-yellow at the distant signal, into a red over yellow at ESS Goodwell and - since the WSS Goodwell was probably still lined for the main at that point, a red at WSS Goodwell.