How can a signal be missed?

[wabash1]

all i am saying is that i can see standard signals 1/2 mile away  clear and approach and on some i can see the restricting and also the stops on others depending on aiming its harder but after a yellow im always going slow so i can stop.. on the LeD signals its about the same just alittle clearer stop signals still look the same.

[blue streak 1]

the only difference was I was looking through two panes of glass not one that the engineer was looking through. I believe the sun was causing some of those sight problems but the angle was the same on that streach of straight track. The weather was clear and visibility unlimited.

[CSSHEGEWISCH]

I don't know how things are elsewhere, but I have rarely to never seen derails as part of the protection at an interlocking except at movable bridges.

[wabash1]

Rode ACE today san Jose - Stockton. Sunny day outside Tracy encountered a SP tri signal showing approach and at about 600 ft spotted yellow signal as the engineer slowed at that time approach speed. Next signal about two miles showed stop which I could only tell about 500 ft from signal. Stopped and proceeded at restricted speed, think I saw code line down causing red signal. Next signals green- new LED type which could tell 2500 ft away. Take it for what its worth guys.  

so your saying that you saw the led signal a half mile away? what was wrong with the other signal that you only saw it at 500 ft away or in railroad terms 10 cars away,  seems to me if you only saw the signal indication at 500 ft ( 10 cars) then you didnt even see the signal mast or signal box or even the area this signal was protecting.

[DMUinCT]

   Zordoz,

   Your right, if the "Derailer" was still in place the Conrail Diesels would have been put on the ground and the wreck would have been avoided.   As it was, just a Penalty Stop was made and the engines drifted into the Amtrak Main.

    Today in the Northeast Corridor, Double Red on a Draw Bridge Approach, the "derailer" is done by opening up the rail.     Note the right rail as photographed from the cab of an Acela.

   

[Jack_S]

The Metrolink train ran at least two signals in order to end up in this final tragic position.

Mark

 Sorry, but I believe you are wrong on this point.  Based on a detailed map (LA Times, Monday  15 Sept 2008) of the situation, which I have posted about in another message in this thread, he only ran one signal.  I will describe the signal situation again below.

The tracks in the vicinity of the Chatsworth Metrolink station are in a sort of an S-shape.  I am writing from the point of view of a northbound Metrolink train, which would come into the area on a West-North-West heading, on a single track main. 

Just before De Soto AVe. there is a south-facing signal to the right (north) of the single track.  Just after De Soto is the south switch for the siding, which is to the main track's left.  Just past the switch there are 2 signals, 1 on each side of the tracks and facing north. 

Just after the switch, the 2 lines curve north and, after about 1 mile, they are headed due north just south of Lassen St..  Just before Lassen there are 2 signals, one on each side of the tracks, both bi-directional, facing BOTH north and south. 

About 1/4 mile past Lassen there are 2 more  bi-directional signals, also on both sides of the tracks.  This 3rd set of signals is just BEFORE the Chatsworth station.  The station is on the right (east) side of the tracks.

About 1 mile north of the station, is the switch controlling the north end of the siding where it rejoins the single main line.  On the right side (east) of the switch is a south-facing signal.  On the left side (west) of the switch is a north-facing signal. 

About 1/8 mile past these signals and the switch, at Rinaldi St., there is a north-facing signal to the right (east) of the now single track.  The rather sharp curve to the west starts right at this signal and, as the investigation showed, visibility around the curve is quite limited.  It seems to me that this last signal is placed to give a south-bound train a preview of the signal to the west of the switch controlling the north end of the siding.  At any rate it didn't play a part in this accident.

So, after passing the signals at De Soto and Lassen, train 111 came to the station with bi-directional signals on each side of the tracks just before the station.  Obviously, under the situation as it existed then, this signal should have been yellow, allowing 111 to proceed into the station to discharge and receive passengers and wait while the freight proceeded safely down the siding.  And, indeed, according to several witness reports in the LA Times, this was the usual procedure.

After the stop at Chatsworth, the train accelerated to 54 MPH, as if the engineer had forgotten that he had just passed a yellow signal.  Fatigue may have been a factor here, since it can do funny things to your short-term memory.  Distraction by text messaging might have been an issue too, even if he did that only while sitting in a stopped train in the station.

So the only signal he blew was the one at the north switch.  His entry into the station was entirely proper.  After that there was only 1 signal, the one just before the crash.  I question the placement of the signals just south of the station.  Having 2 sets of signals, 1 at Lassen, 1 at the south end of the station 1/4 mile away, seems unnecessarliy redundant.  Wouldn't it be better to have them just north of the station, so that the engineer has them sitting in his view throughout any stops? 

In any case, as soon as the engineer accelerated hard out of Chatsworth, his fate was sealed.  Only a quick recognition, well in advance, of the red signal at the north switch could have saved them.  If he went past the switch by 5 feet and stopped, the freight would have still hit Train 111.  Heavy and headed downhill, the UP crew would have been helpless to prevent it.

Jack 

[Railway Man]

I think you meant "Method of Operation" or "train-control system," as TWC and DTC are not signaling systems.

Depends on how many trains you want to run, and if you define "efficient" as "delivers most value for least cost of capital, operation, and maintenance."

For low-train-density railways, TWC and DTC are the most efficient.  There is very little real difference in efficiency between the two.

For high-train-density railways, CTC is the most efficient.

ABS is not a Method of Operation or a train-control system; it's a safety overlay on a TWC or ABS system.  There is some increase in efficiency of operation on a single-track railroad and a significant increase on a double-track railroad, compared to TWC or DTC without ABS.

RWM

 

[Awesome!]

We all have our opinions on signals but I have one simple question. what is more efficient signal system?

Track Warrant Control (TWC),Direct Traffic Control (DTC),Automatic Block Signaling (ABS),Centralized Traffic Control (CTC)

[Sunnyland]

It sounds like NTSB concluded the Metro engineer was texting at the time, so he was probably just not paying attention.  People do stuff like this when they are driving their cars and get into accidents often.  We had a bad highway crash in the area a few months ago where a tractor trailer climbed up over stopped cars and he was on the phone.  People were killed or badly injured in that wreck.  We have light rail here in St. Louis, but it runs on its own track, does not share with any freight RR's.  I think that's not a good idea due to heavy freight traffic and there's a lot of difference between the weight of the train engines. I didn't know any light rail ran that way.  It would be nice to have something that would stop a train when they blew past a red signal, but I'm also sure that would be very costly.

[zardoz]

   The NTSB report I refered to was January 1987, Report PB88-916301.  Crash at "Gun Power Falls", Chase MD.  This is the point where the 4 track Corridor comes down to two tracks to cross the river.

   As I remember the information from then, working from memory, (I was working IN Conrail's new HQ building in Philly, but NOT for Conrail) three Conrail Diesels under control of engineer Gates was on a "deadhead" move from Baltimore to Harrisburg.  He was doing 60mph as he passed the Approach signal, the cab signals worked, BUT, the Audio Alarm was taped over, as it was on some Conrail locomotives.

   When he hit the Red & ATS at the point where 4 tracks become 2, the brakes set in a Normal Application that carried all three locomotives out onto the Main short of the bridge.  The "Black Box" read 62mph.  Despite claims by the engineer, NTSB re-ran the site using the "Black Box" data, the locomotives stopped within a few feet of the crash points.   The Amtrak cab signals went Red at 127mph? (17mph too fast for a 14 car train with two AEM7 locomotives, not all cars had Disc Brakes) and was down to 94mph? at impact. (not sure of the exact Amtrak speeds, it was 21 years ago)   PTC "would have" taken control of the Conrail locomotives when it passed the Approach Signal.

   Some places in the northeast, Cab Signals and ATS goes back to the age of Steam.   

IIRC, a little-known factor leading to the above-mentioned incident was the "cost-saving" move to reduce maintenance by removing (just a few weeks before the wreck) the AUTOMATIC derail from the junction.  Had the derail not been removed, it is likely that the Conrail locomotives would have not made it into the path of the Metroliner.

If you go to the NTSB site http://www.ntsb.gov/ you will see that the NTSB's "Most Wanted" list has one item: Positive Train Control http://www.ntsb.gov/Recs/mostwanted/positive_train.htm

 

[DMUinCT]

   The NTSB report I refered to was January 1987, Report PB88-916301.  Crash at "Gun Power Falls", Chase MD.  This is the point where the 4 track Corridor comes down to two tracks to cross the river.

   As I remember the information from then, working from memory, (I was working IN Conrail's new HQ building in Philly, but NOT for Conrail) three Conrail Diesels under control of engineer Gates was on a "deadhead" move from Baltimore to Harrisburg.  He was doing 60mph as he passed the Approach signal, the cab signals worked, BUT, the Audio Alarm was taped over, as it was on some Conrail locomotives.

   When he hit the Red & ATS at the point where 4 tracks become 2, the brakes set in a Normal Application that carried all three locomotives out onto the Main short of the bridge.  The "Black Box" read 62mph.  Despite claims by the engineer, NTSB re-ran the site using the "Black Box" data, the locomotives stopped within a few feet of the crash points.   The Amtrak cab signals went Red at 127mph? (17mph too fast for a 14 car train with two AEM7 locomotives, not all cars had Disc Brakes) and was down to 94mph? at impact. (not sure of the exact Amtrak speeds, it was 21 years ago)   PTC "would have" taken control of the Conrail locomotives when it passed the Approach Signal.

   Some places in the northeast, Cab Signals and ATS goes back to the age of Steam.   

[CleveUnionTerm]

Whenever men and machines get together, there will be accidents......it goes

back to the Garden of Eden.

[SFbrkmn]

The signal awareness form in my opinion is just more loaded paperwork put on train crews but its a rule that the carriers are enforcing to the dotted line. In the past it used to be a verbal warning if a crew were caught without filling one out, now its automatic investigation w/ level S being handed out to the entire crew--not just the condr. This is in itself a possible distraction in keeping one of these updated while booging down the trk but for jobs w/ just one man in the cab, can't be fun. Amtrk 3&4 runs across the La Junta Sub from La Junta-Dodge City w/ just a lone engr & it is required by him to keep a SAF per BNSF operating rules. Its a rule, bottomline. Most of us don't like it but its not worth getting in trouble over.

[mphill66]

Does anyone know if Mertolink requires its engrs to fill out a signal awareness form while working? This has become standard on freight carriers and they very strict to ensure condrs keep one filled out and turn it in at the end of a shift.

For certain, Metrolink engineers are required to call out signals as they approach them. I haven't ever seen one go by with a pen in hand and as stated above, they have plenty do keep aware of without writing  something every few minutes. Remember he/she is alone in the cab.

[TH&B]

Does anyone know if Mertolink requires its engrs to fill out a signal awareness form while working? This has become standard on freight carriers and they very strict to ensure condrs keep one filled out and turn it in at the end of a shift.

How is writing each signal indication down gonna make you not miss a signal? What if you miss a signal drop in front of you while writing it down? What about busy junctions with several closely spaced signals? What if what you wrote down was not what you realy saw? All it does is document with the same falibal accuracy as any human. Of course it's a cheap solution , but not a solution.

Maybe car drivers should also write down traffic signal indications at intersections they approach. If these are the kind of safety rules US train operators work with then I won't spend any money on buying a train ticket.

Use some technoligy in this day and age for safety's sake like all other modern countries that run passenger trains .

[awalker1829]

I would also point out that unlike automotive traffic lights, the signal lenses employed in railroad signals tend to focus the light into a narrow beam in order to project it further. The net effect is that some signals are easier to miss if you're not looking for them at the point at which the signal is aimed-specifically on sharp curves. If you miss calling the signal upon first sighting it, you may not be able to regain sight of the aspect until you're almost on top of the signal.

[SFbrkmn]

Does anyone know if Mertolink requires its engrs to fill out a signal awareness form while working? This has become standard on freight carriers and they very strict to ensure condrs keep one filled out and turn it in at the end of a shift.

[Paul_D_North_Jr]

From the "NTSB investigation findings ?" thread here:

http://cs.trains.com/forums/1534686/ShowPost.aspx

specifically, the post by "ericsp" on 09-16-2008 at 11:48 PM, who provided -

"Here is the NTSB's railroad investigation report page.
http://www.ntsb.gov/publictn/r_acc.htm

for "Publications - Railroad Accidents",

I found the "Railroad Accident Brief - Collision of Two Southeastern Pennsylvania Transportation Authority Trains - Abington, Pennsylvania - July 1, 2006", NTSB/RAB-08/03 at:

http://www.ntsb.gov/publictn/2008/RAB0803.htm

SEPTA is Philadelphia's mass-transit agency, and these were "heavy-rail" electrified (11,000 v. AC, 25 hertz) multiple-unit commuter trains, operating on one of the former Reading Railroad's commuter lines:

In the context, 3 points are worth noting - if you want all the details, see the full report, it's only about 6 pages:

1.  The "violating" engineer - who is reported to have been trained and qualified, but was only on his 5th solo trip:

"The southbound train engineer did not comply with a total of three wayside signal indications - approach (yellow over red over red), stop (red), and stop and proceed (red) - leading up to the collision.  Note that: "two of which were red, indicating the need for him to stop or stop and proceed."] He also operated through and damaged the siding switch at Grove South, which was aligned for the northbound train to enter the siding. Yet, as the southbound train continued traveling between Grove South and the eventual accident site, the engineer passed three crossings and properly used the train horn at each, and he made two passenger station stops." [Emphasis added]

The report really doesn't indicate why or how he ran past those signals, while still having the presence of mind to make the station stops and sound the horn, etc.

2.  The dispatcher surely didn't cover himself with glory, either, when an alarm indicating a run-past signal sounded and displayed (immediately below), or when called by the operator of the northbound train (see 3. below):

"The train dispatcher acknowledged a system alarm for an overrun signal, via computer mouse click, shortly after the southbound train engineer operated through the siding switch at Grove South. However, he did not take any additional action in response to the alarm. [Emphasis added] The alarm sound and screen text type were not unique, and alarms were a common occurrence for routine matters during a dispatcher's shift. Also, although he was responsible for only one desk/dispatching area, the dispatcher was covering two desks/dispatching areas at the time of the accident because his counterpart was taking a break."

"[I]n this accident, he reflexively cancelled the audible alarm for the overrun signal about 30 seconds after it sounded. Nearly 4 minutes and 20 seconds elapsed between the time when the alarm initially sounded and the collision occurred."

3.  The engineer of the northbound train was pretty sharp, and came close to being a hero, in my opinion, had her efforts in preventing the collision succeeded - she surely tried as hard as anyone could expect:

"A few minutes prior to the collision, the engineer of the northbound train passed a sequence of two signals - a clear (green) followed by a stop and proceed (red) - that concerned her. She was concerned because she had not received an approach (yellow) signal indication before the stop and proceed indication and she was operating the train on single track.  The engineer stopped the train and tried calling the train dispatcher three times to confirm that the signal instructions were correct. After receiving confirmation via radio, the engineer again began to operate the train and proceed in compliance with the signal instructions. When she saw the headlights of the southbound train, the engineer of the northbound train reapplied the brakes, stopped the train just prior to the collision, and told passengers to brace themselves." [Emphasis added]

This sounds similar to the posts by "zardoz" earlier today (09-18-2008) at 9:24 AM (Page 10 of 11, near the bottom) and 10:03 AM (Page 11 of 11, near the top) in the "Commuter & Freight Trains Collide North Of Los Angeles" thread here at:

http://cs.trains.com/forums/10/1532043/ShowPost.aspx#1532043

- Paul North.

[oltmannd]

  Northeast Corridor, yes, Auto Train Stop, and it will stop the train short of the switching point, if the engineer has slowed to approach speed at the previous signal.

   20 yeas ago , at the Crash at Chase MD, the Conrail engines failed slow at the approach signal.   They hit the stop signal at 62mph, the stop inductor tripped, the speed caused the locomotives to drift out onto the Amtrak Main in front of the northbound Patriot.

   Many died, the conrail engineer went to jail.

-------------------------------------------------------------------------------------- 

 CPL!!!  Below is a photo of the right side of an Acela Dashboard.   Note the Cab Signal panel: a CPL Display, Signal Speed (20), and Track speed (- - is 0).  

No stop inductor.  Just plain vanilla cab signals.  Locomotives stopped because the engineer had applied the brakes.

[blue streak 1]

Rode ACE today san Jose - Stockton. Sunny day outside Tracy encountered a SP tri signal showing approach and at about 600 ft spotted yellow signal as the engineer slowed at that time approach speed. Next signal about two miles showed stop which I could only tell about 500 ft from signal. Stopped and proceeded at restricted speed, think I saw code line down causing red signal. Next signals green- new LED type which could tell 2500 ft away. Take it for what its worth guys.  

[mphill66]

....... Accordingto the media, a preliminary report by the NTSB said that he missed the signal.  I beleive the crew was killed in the crash, so we will probably never really know why.

 

Actually, the only crewman killed was the Metrolink engineer. If you noticed in the images of the crash, the ML engine was forced backward into the first passenger car.

The ML conductor and the 3 UP crew members survived.  NTSB was going to interview a UP crewman today... don't know if that happened or if it did whether we'll get to hear any of that info.  Also, they are trying to interview the ML conductor, but that is delayed until he can have a Representative attend the meeting.

[mphill66]

...................

The Metrolink train ran at least two signals in order to end up in this final tragic position. The dispatcher saw the situation, but only manged to reach the conductor by radio moments after the accident had occurred. No automatic train stop on this system. 

Mark

An alert sounded at dispatch but way too late to do anything. NTSB and the RRs ran tests today, with both Metrolink and the UP local trains configured as they were that night.  They determined that the engineers had less than 5 seconds to react once they could 1st see the other train (this was just north of a curve) . The Metrolink engineer did not apply brakes, the UP crew did after a 2 second reaction time, so full emergency braking 2 seconds prior to collision.   The Metrolink was slowing to 42 in anticipation of the tunnel speed limit just ahead. 

[mphill66]

......  If the passenger train was taking the siding to meet the UP train (as is my understanding),...

Karl

CSX Train Dispatcher 

Karl, as you may have heard by now, the Metrolink was to hold on the main and the UP was lined to the siding..

[BaltACD]

As reported, the Metrolink train did run through a switch that was set against it. Apparently hit it at high speed and distorted it in such a way that the train remained on the track. Too bad, as a derail might have been better in this case.

Mark

 

Run through (trailed through) switches do not derail trains....not until they attempt to make a move over the switch facing the point, as the points are now unsecured and can move about at the movement goes across it..

[Phaeton]

As reported, the Metrolink train did run through a switch that was set against it. Apparently hit it at high speed and distorted it in such a way that the train remained on the track. Too bad, as a derail might have been better in this case.

The Metrolink train ran at least two signals in order to end up in this final tragic position. The dispatcher saw the situation, but only manged to reach the conductor by radio moments after the accident had occurred. No automatic train stop on this system. 

 

Mark

 

[albal]

Not only the signal, but how can two trains be on the same track without one of them having gone through switches set against them, without derailing?

[Randy Stahl]

....I suppose we have to assume signals that must be read by humans...will not be 100% since we all can commit errors.

Let me guess .. no one on this forum has EVER run a red traffic light ?

[tree68]

BaltACD - Thanks for fleshing that out.  There is at least one CPL at Deshler (JoeKoh could probably tell you from memory) that has side lights as well.

[BaltACD]

Three shots of the same signal at Deshler (looking south from the diamond):

I forget what the lunar modifier at the top of the mast means.  Somebody with a CSX rulebook can fill us in.

The train for which the signal was set came around the SE wye and headed south.

The top light, is just White, not Lunar and is commonly refered to as the 'Pilot Light'.  The signal can be configured to display a Lunar aspect that is the main unit with Lunar being diagonal to the left.

If the top Red signal had a Pilot Light displayed, it would be a permissive 'Stop & Proceed' indication rather than the fully restrictive 'STOP' of the signal indicated.  An Intermediate signal, with a number plate, displaying a STOP indication, without a pilot light, would be a 'Stop & Proceed' in the old B&O interpertation.

A Green display without the pilot light would be a Slow Clear indication, mandating Slow Speed for the entire train over the interlocking the signal protects....with the pilot light it is Clear - proceed at maximum authorized speed.

The B&O CPL's, could have pilot lights in all six areas about the signal.  Top, Bottom, Top Right, Top Left, Bottom Right and Bottom Left .... each identified a speed modification to the basic color of the signal.  The signal without a pilot light would be the most restrictive of any of the indications for each color displayed.