Tuesday, February 23, 2010

NTSB Report Questions Brake Timing

According to the NTSB's "Rail Collision Vehicle Performance Study," released Jan. 26, it would appear the operator of the colliding train had an opportunity to avoid a collision on June 22, 2009.

As can be seen in figure 10, the case where brakes are applied at the first of partial view of train 214 [the train that was struck] results in a stop short of the point of collision even with minimum brakes.

Figure 11 shows the case where nominal brakes are applied with a 3 second reaction time after the first of partial view of train 214. Again the train would have stopped short of the point of collision.
The report also reveals that "there was no communication between the operators of train 112 [colliding train], the stopped train 214 or the Metrorail operations control center before the collision."

(h/t anonymous tipster)


Anonymous said...

Well, WMATA said she wasn't texting or using a cell phone, but...what a terrible and AVOIDABLE tragedy.

Anonymous said...

The "what if" scenarios should be lengthened to say "what if someone, anyone was paying attention."

Matt' said...

Thanks for posting this.

I would point out that this is just a technical report on the braking characteristics. The full NTSB report will talk about other factors that could have played a role.

For instance, the NTSB report on the striking of wayside workers at Eisenhower Avenue noted that because of track geometry, the operator would not have been able to tell which track the workers were walking on at the first partial view.

The same case may apply in this instance. If the operator of the striking train (112) couldn't tell, she might have thought the stopped train (214) was an oncoming train on the Glenmont track.

Your post did not mention that at the first "full view" of the stopped train, even with full braking, there still would have been a collision - at approximately 24 mph. If a 3 second reaction time was allowed for, the collision would have occurred at about 44 mph.

The technical report which you linked to indicates that the emergency braking occurred at a point where the collision speed would be about 33 mph. This seems to show that the emergency brakes were applied less than 3 seconds after first full view of the stopped train.

Anonymous said...

I confess a lot of what I'm reading here is a wee bit above my head. However, I cannot help but think - yet again here we go potentially blaming a dead person. The dead never sue and never contradict so I suppose it works...?

Unsuck DC Metro said...

Don't think the report IDs any blame. Would be very surprising if the conclusion as to the cause of the collision was solely braking. These things are usually a chain of events

Anonymous said...

A chain of events. "What we have here is... a lack of communication!" (Wonder if the brat pack on the previous post has any inkling of where that line comes from. hah!)

TimK said...

I would like to know two things.

(1) Why had the operator of train 214 not communicate with OCC? Had 214 just stopped immediately before it was struck?
(2) How do WMATA's train radios work? Can a train operator on a given line hear other trains' communications with OCC?

In this case, had 214 communicated with OCC and had the operator of train 112 been able to hear that communication, she would have known that there was a stopped train ahead of her and been able to watch for it and avoid the collision.

Anonymous said...

Going off on a slight tangent...did anyone read the interview with the operator of the train that was crashed into? At one point he mentions how the pocket track at Farragut North (where the recent derailment was) had issues over a year prior to the crash last June. That would mean that there were problems there for almost two years prior to the recent derailment. Guess that is the reason why no one would tell the board anything about why the derailment may have happened.

Anonymous said...

1:20 Anonymous is right...that operator interview is just frightening...try reading that and then NOT being worried about who's driving the trains, and how safety issues are handled.

Anonymous said...

There is a lot of material to read. Any errors are due to my memory.

For TimK:
- 214 had a train visible ahead of it. The operator assumed the loss of speed commands (e.g. stopping) was due to the train in front.
- In general, everyone can hear OCC and OCC should be able to hear anyone. You will only hear a train or a portable if you are nearby.

Frankly, I think the OCC part of the report is the scariest part I have read. The language is thick, but basically train 214 became sort-of invisible to ROCS (rail control computer) before the accident, and the OCC staff really didn't get a lot of useful warning.

(this represents my opinion, not my employer, etc)

Anonymous said...

You very likely won't hear from the front line technicians at the "public hearing"
Their interviews with NTSB are behind closed doors, with Metro lawyers galore debriefing them.

Anonymous said...

This is what I said at the time: why were the brakes not applied? And there was a ton of dialogue to put it politely with commenters some said she did all she could, others saying, well let's wait on that one. And suffering the abusive language for daring to say that.

Anonymous said...

I thought the faragut derailment happened to prevent the train going on the wrong track.

Anonymous said...

These documents provide a little more context on the sight tests referenced in this study:

Aerial photo annotated with the sight test locations:

Photos from the operator's perspective at the various sighting distances:

Anonymous said...

For anonymous 2/23 9:15 - There appears to be physical evidence that the brakes were applied. If nothing else, the emergency stop button (aka The Mushroom) in the striking train was depressed when recovered (The console was mostly intact). I believe they also have some skid mark evidence from the rails. I'm still trying to digest all the NTSB stuff.

For anonymous 2/23 9:16 - The short answer to your comment is "Yes, that is what happened."

Pocket tracks are used for reversing trains. They are designed to accept trains from either main track in either direction. Here is a picture: http://commons.wikimedia.org/wiki/File:WMATA_D_Route_Pocket_Track.jpg

The pocket track at F North had been rigged so it could only be used by trains approaching/leaving from the south (towards F North station). The north side of the pocket track had a derail installed to prevent a train passing onto a main line track from the north side of the pocket. Here is a bit more on derails http://en.wikipedia.org/wiki/Derail_%28railroad%29

The derail rig was intentional as a safety measure. What is unclear (to me) is whether or not the derail was a simple metal wedge (installed manually) or an automatic safety device. I know what has been said (automatic), but I'll wait for some concrete data.

The questions to ask are:

a) The train was outbound to Shady Grove. It should have stayed on the main line. Why did it divert to the pocket track? Changing the switch (interlocking) is typically done at the control center.
b) Why did the operator continue to proceed north when it became clear he was on the pocket track?
c) Was the train in Automatic or Manual? Even if it was in Manual, why didn't some of the automatic systems intervene?

In case you are still a bit foggy, here are a couple of pocket track locations you can eyeball for reference...

Red Line: Between Silver Spring and Forest Glen
Orange: West Falls Church station, Between Stadium/Armory and MN Ave
Yellow: National Airport
Green: None on Surface/Above Ground
Blue: Between Stadium/Armory and Benning Road, National Airport

(this represents my opinion, not my employer, etc)

Anonymous said...

A couple of additional bits about F North. I read some of Matt Johnson's (@tracktwentynine on Twitter) excellent stuff and can now ponder on some of my earlier comments.

a) The derail was automatic in nature, as opposed to a derail which would be clamped on a rail and left there. Frankly, I prefer the New York system where if you hit a Trip (activated by a red signal) the train stops by an emergency application of the brakes.
b) The train may have been routed onto the pocket track because of a confusion over the destination of the train. Apparently that train was intended to turn around at F North (at one point) and was then changed to turn around at White Flint.

(this represents my opinion, not my employer, etc)

Anonymous said...

I think there is something inherently wrong about focusing on metro's operators, how they suck, and not more on the management and leadership of metro who oversee these employees. Who hires the people? Who is in charge of training and monitoring performance? Is there any internal system of holding people accountable at metro, especially those in leadership positions?

All of these problems - funding issues, broken down equipment, lack of focus on safety, employee performance - are just symptoms of the much bigger organizational problems metro faces.

I'm not saying that metro employees should not be held accountable for their actions, but the bulk of the responsibility lies with metro leadership.

I've worked in dictatorship management environments like metro has and frequently what happens is that employees see issues, make recommendations on how things could be improved, they get ignored or told no repeatedly, aren't given any good explanations for why or viable alternatives, and over time, they stop caring entirely because it doesn't make a difference to care.

Metros biggest problem is that they have bred a culture of indifference that will be very difficult to overcome.

Anonymous said...

Have you seen the pictures of the recreation simulation? It would be unreasonable for the operator to brake as soon as the "partial view" was achieved, I can't even tell it's a train in the distance, much less which track it's on (it's around a bend). I think the driver was not in error and the evidence suggests that they applied brakes soon after the stationary train came into full view.

Post a Comment

Creative Commons License
This work is licensed under a
Creative Commons Attribution-Noncommercial 3.0 Unported License.
Site Meter