Could The Operator Have Seen The Train?
In reading coverage of the metro accident yesterday, I was pretty surprised to see people -- both official statements, and comments from the usual lynch mob who know absolutely nothing about the circumstances of the accident -- quick to place blame on the operator of the moving train.
An article yesterday is titled Experts Suspect Failure of Signal System, Operator Error. Commenters everywhere quickly blame the train operator, herself a victim of the accident, without having all the information. The investigation is ongoing, but there is physical information available to everyone about this crash that is being overlooked.
I have ridden that stretch of track many times since I work in Silver Spring. It is above ground, and a lot of the trip is pretty wide open. Visibility should be good. But I remembered that there are several curves and that the train frequently travels close to trees and buildings. I wondered to myself, is it possible that the accident could have taken place in a scenario where the operator had very little visible warning of the stopped train ahead of her?
The image above is from Google Maps. Coincidentally, the satellite happened to catch a metro train traveling pretty much where the second train was moments before the crash. As you can see, there is heavy tree cover west of the tracks around New Hampshire Avenue, limiting the line of sight around the corner. The bridge itself also surely presents a significant visual obstruction.
The red line shows roughly where a train operator could see. Anything to the right of the line would be out of sight from the position of the train in the picture. According to the scale in the picture, the distance between the train and the bridge is barely 400 feet. So if the second train was stopped a couple hundred feet on the other side of the bridge, just outside the visible line, the operator couldn't even have seen the train until she was within about 500 feet of it!
The speed limit on this part of the track was apparently 59 miles per hour. This is 87 feet per second. Supposing she could only see the train when it was within 500 feet of her, she would have less than 6 seconds to react and bring her train to a complete stop to avoid a collision.
I don't have any idea what the stopping distance for an 8 car metro train traveling at about 60 miles per hour would be. But for a car, it's about 185 feet, and for a truck, it's about 270 feet, according to NHTSA. One must imagine that for a metro train, which has does not have the benefit of rubber tires (steel wheels on steel tracks have a lot less friction than rubber on asphalt) and is also a hell of a lot bigger, it is significantly longer.
It seems possible, even likely, that in the right set of circumstances, this accident could not have been avoided by the operator even if she had reacted the moment she saw the train ahead of her.
In The Grassy Knoll
So where exactly was the stationary train? In all the coverage yesterday, I could not find a photograph of the crash site taken from overhead to confirm it's position relative to the bridge. However, The Post published a graphic showing the logistics of the accident, shown to the left.
I have to say, this looks pretty damn incriminating for a train in the blind spot. Compare to the graphic above. If this is accurate, the operator probably could not see the train until too late.
I read this morning that investigators "plan to experiment with similar trains to determine approximate speed and stopping distance." Frankly, I'm pretty surprised that the stopping distance of a metro train (given length and number of passengers) wouldn't be well known to people in the industry. You'd think they would have some idea when a typical train needs to start slowing down. But then again, I'm also pretty surprised that nobody else has mentioned the possibility that maybe, the operator didn't see the other train until it was too late.
Obivously, there was a computer malfunction that was the primary cause of this accident. This situation should never have happened in the first place. Once the failsafes were gone, only the train operator could have stopped the trains from colliding. But if she couldn't see the other train until there was five seconds and 60 miles per hour between them, there may have been nothing she could do. It has been reported that the emergency brake had been activated at the time of the crash, but they can't say for sure if it worked. Maybe there just wasn't time to stop the train.
7 comments:
SamR here. NYCT Train Operator. From my understanding, similar type trains here in NYC have a braking power of about 2.5m/s/s. If you assume that the train was going 60mph - about 30 m/s, then the train would come to a stop in about 180m or 550 ft or so. (this sounds about right to me - I often enter a station at around 40mph, and it takes about 1/2 the station to come to a stop at a full service brake - no that I brake that way - but that is about 300ft), so at 60mph, the distance would be significantly more, due to the v-squared component in the Kinetic energy of the train, which is the relevant physical quantity here.
combined with your topographic analysis - seems pretty clear that the T/O didn't have a chance. gotta love Robo-trains. The bosses here in NYC are slobbering trying to get them put in throughout the whole system.
SamR
SamR again. I forgot one other, very important factor (and this may be why the officials need to check the braking distance): the grade of the track, i.e. if there is a hill of any sort. Obviously, if there is a downgrade, then the braking distance increase, while if there is an upgrade, it decreases. A minor grade can have a major influence on the stopping distance.
sam
According to the NTSB report on the Shady Grove Collision in 1996, the maximum braking power required by Metro's specifications is 3.2 mph/sec during an emergency application.
I did the calculations, but from the aerial pictures and videos, I was able to place the rear of the stopped train at approx. the small building just S of the bridge.
This leaves an approx. line of sight of close to 2100 ft.
Using an avg. motion of 58ft/sec., she theoretically would have had around 30 seconds to react and stop. She hit the brakes 400 ft. before impact. Maybe 7 seconds.
What was she doing for the other 23 seconds???
Her nails?
Chris - that just can't be the case. There is no way the operator could have seen the train from even 1000 feet back.
Look at this picture showing the line of sight from several spots further back, and placing the target exactly where you say. She would have to have been able to see through trees and walls.
http://www.flickr.com/photos/jamietre/3662751852/sizes/o/
To make this even more clear, here is a google street view looking north from the bridge on New Hampshire Avenue. And this is an even better vantage point than the operator would have had because it's from above. You clearly cannot see even a little bit around the corner.
http://www.flickr.com/photos/jamietre/3661962647/sizes/o/
Thanks to Matt for the info about emergency braking rates. If the speed before collision was 59 mph, then the stopping distance at 3.2 mph per second would be about 800 ft. If the brakes acted with 300 ft to go, speed at collision would be about 46 mph, and the time to collision would be about 4.0 sec. Passengers enduring this event might be unable to remember the train slowing before impact.
Robo trains need to be put in as soon as possible to limit the error rate for stopping distance for NYC trains.
Post a Comment