WMATA Red Line Collision

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Wow, this mass transit sure is dangerous.
Yeah, not so much. How many people died on the roads today?
IMHO, it is how people view such accidents.

Just how many highway accidents, which involve only two vehicles, result in 7+ fatalities (this Metro accident)?

Just how many highway accidents, which involve only one vehicle, result in 250+ fatalities (Air France)?

In other words, there are fewer accidents on mass transit, but when they do occur, many more lives are lost in that one incident.
 
Speaking of the 1000 series cars, the NTSB has already said in a press conference today that it recommended to the WMTA and the builder of the cars that they needed more structural strength or that they should be retired otherwise. Neither was done apparently and the NTSB's predictions of what would happen to the cars in a crash are now quite visible in the accident photos. Sadly. :(
 
Speaking of the 1000 series cars, the NTSB has already said in a press conference today that it recommended to the WMTA and the builder of the cars that they needed more structural strength or that they should be retired otherwise. Neither was done apparently and the NTSB's predictions of what would happen to the cars in a crash are now quite visible in the accident photos. Sadly. :(
Yeah, the NTSB sure didn't pull any punches:

NTSB’s Debbie Hersman this morning confirms that the the striking train was a 1000-series car and that the struck train was a mix of 3000- and 5000-series. She notes that the NTSB has "long been on record" about the crashworthiness of the 1000 series. "We recommended to WMATA to either retrofit those cars or phase them out of service," she says. "Those concerns were not addressed."
 
The DC system is about the most automated system in the US. In fact, whereas most US rail transit systems are operator controlled and computer/automation backed, the DC system is reverse - fully automated with human back up. This arrangement has been a problem with WMATA for years.

Somewhere in the over 3000 articles about this incident was a note that the operator on the rear train apparently engaged the manual override but for whatever reason, it didn't engage. Could be an overzealous reporter for all I know, because as mentioned, the 1000 series doesn't have a data or voice recorder.

With regards to AlanB's noting of the NTSB comments, sure, it may have lacked structural strength, but again, if the accident were avoided, it would be a mute point. No matter how structurally sound a vehicle is, the g forces on a body decelerating from 40 or 50 MPH to 0 in half a second can damage any body. Don't forget one of the Metrolink disasters from a few years ago where the train hit a Jeep at a grade crossing, the fatalities came from passengers receiving blunt force trama to their abdomen from the club seating table arrangement.

To answer Murjax, I don't know. I have heard of operators asking to manually override the system and being refused, but I can't really say if ALL the operations are automated. However, I do assume that there is a command center, but that the switches and signals are automated.
 
Does the Metro have dispatchers, or someone in a command center controling the switches and signals, or is that automated too?
I assume that they have every kind of control available as the system is supposed to be state of the art. That means there should have been some form of automatic train stop if the operator is incapacitated or if the train ignores a speed restriction or red light and I wonder if they don't have GPS tracking plus they have a huge dispatching center where they know where every train is at every second. They run thousands of trains everyday and they don't run into each other. So something extraordinary happenned to make this accident occur. If you notice, it happenned on a blind curve. So I just wonder if the operator had somehow taken manual control became frustrated with the delay and ignored a signal or speed restriction. The only other possibility that I can think of is somehow the trains computer went nuts or it's systems failed somehow. That they can probably determine even with the devastation. Think there might be a few hundred lawsuits after this??????????

This from wikipedia:

On January 6, 1996, during the Blizzard of 1996, a Metro operator was killed when a train failed to come to a stop at Shady Grove station. The four-car train overran the station platform and struck an unoccupied train that was awaiting assignment. The National Transportation Safety Board (NTSB) investigation found that the crash was a result of a failure in the train's computer-controlled braking system. The NTSB recommended that Metro grant train operators the ability to manually control the braking system, even in inclement weather. Additionally, investigators recommended that Metro prohibit parked rail cars on tracks used by inbound trains.[33]

The November 3, 2004 accident at Woodley Park-Zoo/Adams Morgan station.On November 3, 2004, an out-of-service Red Line train rolled backwards into the Woodley Park-Zoo/Adams Morgan station and hit an in-service train stopped at the platform. No one was killed, but 20 people were injured.[34] A 14-month investigation concluded that the train operator was most likely not alert as the train rolled backwards into the station. Safety officials estimated that had the train been full, at least 79 would have died. The train operator was dismissed and Metro officials agreed to add rollback protection to more than 300 rail cars.[35]

During normal operation on revenue tracks (used for passenger services), trains are controlled by an automatic train operation system that accelerates and brakes the trains automatically without operator intervention. However, all trains are manned with train operators who close the doors (they can optionally be set to open automatically), make station announcements, and supervise their trains. The operator can manually override a train when necessary.[24]

http://en.wikipedia.org/wiki/Washington_Metro

and:

http://www.wmata.com/about_metro/board_of_...ionStopping.pdf

http://en.wikipedia.org/wiki/Automatic_train_operation
 
Last edited by a moderator:
There are two types of manual operation. One, the train operation system (ATO) is taken over by the operator, but speed compliance and safety is still ensured by Automatic Train Control (ATC). The operator cannot force the train over the speed limit for the location and cannot operate if the block is red.

The second type of manual operation is full manual with all systems disengaged - ATO and ATC. This type of operation is limited to a very restricted speed - typically 15mph. It is used rarely, and only under the direction of the dispatcher.

The automated operation of Metro is similar to BART in the Bay Area and PATCO in South Jersey.
 
There are two types of manual operation. One, the train operation system (ATO) is taken over by the operator, but speed compliance and safety is still ensured by Automatic Train Control (ATC). The operator cannot force the train over the speed limit for the location and cannot operate if the block is red.
The second type of manual operation is full manual with all systems disengaged - ATO and ATC. This type of operation is limited to a very restricted speed - typically 15mph. It is used rarely, and only under the direction of the dispatcher.

The automated operation of Metro is similar to BART in the Bay Area and PATCO in South Jersey.
This from "All Aboard":

My first impression is that the operator of the second train was creeping up on the first one. This is standard procedure on rapid transit systems, not usually a problem unless the operator violates procedures.

 

Automatic block signals allow trains to proceed after coming to a full stop, except for those protecting interlockings. On New York City subways, you can see the tripper drop after the train stops. This allows the train to pass the signal. After stopping, the train is supposed to proceed at restricted speed, able to stop within half the distance that the operator can see.

 

In this case, the operator passed two signals this way, then accelerated far beyond restricted speed.

 

The Long Island Rail Road had a terrible rear-ender west of Jamaica in the early 1950's. The operator of the second train saw a signal ahead change from stop to clear, but it was for the train between him and the signal!
 
IMHO, it is how people view such accidents.
In other words, there are fewer accidents on mass transit, but when they do occur, many more lives are lost in that one incident.
For auto accidents news stations should do a best of the week. They will surely find more than nine deaths and 70 injuries.
 
Yeah, the NTSB sure didn't pull any punches:
NTSB’s Debbie Hersman this morning confirms that the the striking train was a 1000-series car and that the struck train was a mix of 3000- and 5000-series. She notes that the NTSB has "long been on record" about the crashworthiness of the 1000 series. "We recommended to WMATA to either retrofit those cars or phase them out of service," she says. "Those concerns were not addressed."
The recommendation that the 1000 series trains be retired long ago, is apparently the lead story this evening for the 5pm TV news.

However, as it appears now, the second train being a 1000 series is only applicable to how it survived (or not survived) the crash, not that it being a 1000 series caused the crash, right?
 
Last edited by a moderator:
This from "All Aboard":
My first impression is that the operator of the second train was creeping up on the first one. This is standard procedure on rapid transit systems, not usually a problem unless the operator violates procedures.

 

Automatic block signals allow trains to proceed after coming to a full stop, except for those protecting interlockings. On New York City subways, you can see the tripper drop after the train stops. This allows the train to pass the signal. After stopping, the train is supposed to proceed at restricted speed, able to stop within half the distance that the operator can see.

 

In this case, the operator passed two signals this way, then accelerated far beyond restricted speed.

 

The Long Island Rail Road had a terrible rear-ender west of Jamaica in the early 1950's. The operator of the second train saw a signal ahead change from stop to clear, but it was for the train between him and the signal!
On an automated system in manual operation (ATO cut-out), the operator cannot violate the speed restriction even if he or she tries. The ATC will not permit it even in manual operation. If ATC says 20mph, the operator can push the throttle all the way forward if he wants. Once the train hits 20mph, the throttle will drop-out and the brakes will engage. If the ATC says stop, the train stops, no matter what the operator does. Total manual operation with ATC cut-out is used very rarely, only under the direction of the dispatchers, and never at more than 15mph regardless of what the signals say. I believe train speed control limits the operation to 15mph whenever ATC is cut-out.

The Washington Metro is not the LIRR of the 1950's. It is a modern, automated rapid transit system. The train control system limits what the operator can do. Once again, I hesitate to speculate about cause, but just about the only viable scenario for this type of accident on a system controlled by modern ATC is a failure of some aspect of ATC system. The operator simply does not have enough control to blunder into this accident.
 
A few thoughts, in no particular order:

In the early days WMATA was way too enamoured with automation, and I suspect still are. Some of it is not too bright. The fatal collision at Shady Grove a couple years ago was a train under automatic control being operated on icey rail. The operator wanted to go manual so he could take more distance to stop, but was told NO by the control center. He had a choice, lose his job or risk his life. He bet wrong and lost.

The WMATA cotrol center was, and maybe still is inadequately small. Teh assumption was that automation would make the hand of man near unnecessary. Were they ever wrong, and still do not seem to have learned.

If Rhor built cars, these are the first set of cars. They had a lot of teething tourbles. I think these were the first rail cars Rohr had ever built, and maybe also the last.

The WMATA cars are all lightly built. Strength can be had without excessive weight, but the car builders in Europe and elsewhere propogandize that such is not possible. Management buys into that so does not require the strength that they ought to.

When the analysis is done, it will be found out that there were multiple errors and defects that all came together. There will be no one single cause.

Fail-safe failed to happen.

The "derailment in 1982" that was mentioned occurred because the control center threw a switch under a train (which should not be possible with normal track circuits), and actually under a car, so that one end went down one track and one end went down the other, crusing the side of the car against the middle wall between tracks that began just beyond the crossover. This accident was further compounded by the fire department finding that their radios did not work in the Metro tunnels. There was a "leaky coaxial cable" that would pick up the Metro frequencies, but not the fire department frequencies. That I regard as beyond stupid.
 
A few thoughts, in no particular order:

The WMATA cars are all lightly built. Strength can be had without excessive weight, but the car builders in Europe and elsewhere propogandize that such is not possible. Management buys into that so does not require the strength that they ought to.
Still banging that old drum? Best way to survive a crash is not have one in the first place. Anything else is just window dressing.
 
A few thoughts, in no particular order:

The WMATA cars are all lightly built. Strength can be had without excessive weight, but the car builders in Europe and elsewhere propogandize that such is not possible. Management buys into that so does not require the strength that they ought to.
Still banging that old drum? Best way to survive a crash is not have one in the first place. Anything else is just window dressing.
To be fair to George Harris, it appears WMATA did everything they realistically could to eliminate crashes from happening (automatic train stop, automated controls to virtually eliminate the human error element, etc.), but as George said, it's foolhardy to assume that an accident will NEVER happen despite every possible effort to prevent it. There is something to be said about build quality and crash standards as a last-resort backup.
 
It appears that there was a "perfect storm" of problems—with braking, signaling, train-operator procedures, car-body construction, etc. However, until everything is known (and we hope all the problems will eventually be identified and corrected), it is useless to speculate as to what exactly caused the accident.

I'd still rather be in a train wreck than a plane or even a car accident. In fact, I was once in a train accident, on the EB Empire Builder in June of 1973 when it struck a cultivator being pulled by a tractor a few miles west of Minot, ND. AFAIK, there were no injuries, though the train was delayed for a couple of hours once it got into Minot.
 
There is something to be said about build quality and crash standards as a last-resort backup.
Doesn't that reduce the amount of energy that is saved taking mass transit?
Somewhat, but not entirely, especially if technologies like regenerative braking are used. No matter how you slice it, steel-wheel-on-steel-rail will almost always beat rubber-tire-on-asphalt (especially single-occupancy-vehicle-rubber-tire-on-asphalt).

I'm not saying I'm 100% supportive of full FRA crashworthiness standards and the heavy trains in America (though big, heavy trains are fun to look at...make the little European ones look like toys! :p ). I'm just saying there's something to be said for the resiliency of heavier equipment in the event of a protective system failure like Metro just experienced.
 
...If Rhor built cars, these are the first set of cars. They had a lot of teething tourbles. I think these were the first rail cars Rohr had ever built, and maybe also the last.
Rohr built the first generation BART cars, which entered service in 1972. Metro started in 1976. The Rohr Metro cars are similar in many ways to the first BART cars. These cars where built in the recession era of the early 1970's when aerospace was in the dumps and companies like Rohr were trying to find some other market. BART offered that opportunity, and the fact that Rohr was a California firm helped a bit was well. Once Rohr had the tooling and the production set-up for BART, the Metro work was a natural follow-up.

In all fairness, after in initial shake-out, the Rohr cars have been pretty reliable at both Metro and BART. The first generation PATCO cars (southern NJ) had their problems as well, and they were built by Budd. I think that any new designs that push the technology envelope have start-up issues.
 
Last edited by a moderator:
True, but to ignore the fact that sometimes accidents happen would be foolhardy.
Yes, but its a subway train, how heavy do you make it exactly? Lets face it, running into the back of another train is a very very silly thing to do, FRA regs are of little consequence then, when you are bouncing round in the car.

That sort of rear ending crash is just not a good thing, same as a head on. All bets are off by then.

I will never agree with the approach in the US of making trains crashworthy, as it is the classic case of door locking after horse bolting.

Sure, you make your cars as strong as you can, but your prime interest should be to keep the trains apart as far as possible. That's more achievable than the crash proof train, nearly as daft as the crash proof plane or the unsinkable ship......
 
True, but to ignore the fact that sometimes accidents happen would be foolhardy.
Yes, but its a subway train, how heavy do you make it exactly?
Heavy enough to not disintegrate the way the 1k car did on impact. The idea that we'll be able to fully prevent railcars from ever coming into contact with one another is the very definition of hubris.

Look, I understand US v Euro crashworthiness standards is a matter of deeply held debate that nobody is ever going to change their opinion on. So perhaps rather than sully this thread with a back and forth argument that accomplishes little we can leave this matter be for the time being?
 
Back
Top