Red Line Collision

TallEddie_81 wrote:What type of train control system is in place to prevent this from happening? I know that the Metro trains are controlled by computer and essentially operate by themselves, and the operator pretty much just operates the doors, unless the train is in a manual mode.

From reading this forum I know that freight trains can pass a stop signal, and still be able to proceed, albeit at a restricted speed. How do the safety systems on freight rails compare to the WMATA system.

Metro does have an ATP (which I know very little about), but the trains apparently were operating in manual mode at the time.

Here is continuing live coverage of the Washington Metro accident:

http://www.nbcwashington.com/

http://www.wjla.com/

http://www.wusa9.com/

http://www.washingtontimes.com/

http://www.washingtonpost.com/

Local Television out of DC just announced that the death toll from the Metro accident is now 6.

Live streaming coverage is also available at:

http://www.wjfkfm.com/

TallEddie_81 wrote:What type of train control system is in place to prevent this from happening?

I was wondering the same thing. In the NYC subway, every signal is accompanied a "train stop" -- the system is explained in great detail here -- but in a nutshell, if a train passes by a red signal, the train stop, a small metal "T" next to the rails, will engage the train's emergency brakes. The lever moves out of the way if the signal is displaying any aspect that permits movement.

So, even if an engineer happens to drop dead at the control stand and they fall forward in such a way to keep the throttle engaged, the train can still be stopped by passing any red signal -- and there's no way one train could hit another without running at least 2 red signals.

Long story short, according to that page, this technology has been in use since the NYC subway opened, more than 100 years ago. The Washington Metro has only been open since 1976. How/why did Washington manage to overlook such old-fashioned, yet very reliable technology?

What type of train control system is in place to prevent this from happening?

The DC metro uses audio frequency cab signals with positive stop enforcement. There is a central computer that can slow trains further to help headway spacing and merging at junctions. The central computer cannot order an unsafe operation because local cab signals govern. Manual mode disables the automated station stopping and central speed targets, but cab signals still govern. Complete cab signal cutout should limit trains to 15mph and requires absolute block protection

Could there have been some sort of failure in the on-board cab signal system so that the train received a false positive type indication? I know that the central computer is mostly for schedule keeping and the local cab signals handle the vital safety functions, but does the central computer also send its speed commands to slow trains for dynamic conditions? If not then if the local system suffered a Type I error (possibly since it began its run) the train would have continued along without signal indication and ramming into the first dynamic obstacle in its way.

How/why did Washington manage to overlook such old-fashioned, yet very reliable technology?

Perhaps they didn't? I don't recall the formal name for the tripper arms, but they are hardly a timeless, perfect solution. They have shortcomings, as does any approach. ISTM they are one of the factors that prevent/constrain swapping of equipment between the various NYCTA lines. They can freeze up in ice storms/blizzards.

WMATA uses different approaches, ATC, and its parts: ATS, ATO and ATP, to reach the same goals. We'll have to wait a while to learn what went wrong in this case, but to claim at this stage that if only WMATA used X, this wouldn't have happened is rather presumptuous.

How/why did Washington manage to overlook such old-fashioned, yet very reliable technology?

Because they preferred to use cab signals, another old-fashioned, reliable technology that date from the 1920's.

No heavy rail system built since the 50's uses trip stops and many have ripped the existing ones out when the lines were resignaled.

Okay, I worked for NYCTA for five years in the 1980s, and we managed to run 90-second headways with manual trip stops and never killed anybody. WMATA has already killed several people.

I'm not an expert on cab signaling, Jersey Mike. What is a "Type 1" failure? WMATA doesn't allow trains to operate without functioning cab signals. Ride the head end of a train, and you'll see three speed displays. First is the actual speed the train is operating, second is the speed commanded by the central computer, third is the maximum allowable speed in the block (from wayside ATS equipment). Speed commanded by the central computer must always be equal to or less than the speed allowed by the ATS equipment.

This accident appears to have been a full-speed collision, which means the wayside equipment was displaying MAS (70 MPH at that location). With an occupied block ahead of the train, that should have been impossible. A failure of the wayside equipment should cause display of a zero speed command, stopping the train. So what the hell happened?

Pnaw, I would seriously question your line of logic. Because one system has a particular type of safety system such as the T-handle does not mean another system does not have a different system, a system which may have been billed as "better" or "equal" to the T-handle. It may be an electronic system. In my industry, cranes, the secondary mechanical brake was replaced in the 1970's by a secondary regenerative brake, and I'd much rather stand under a dropping load with a modern regen brake than trust a mechanical device. Point being, blanket statements without further investigation is not wise.

Further, most railroads in America, transit or interchange, operate without the T-handle. This includes busy railroads with short headways such as CTA, Metra Electric, Northeast Corridor, and Boston subway.

What we are faced with is a tragic accident and we're going to have a lot of needless finger pointing distracting us from the real cause which remains unknown at this time.

A failure of the wayside equipment should cause display of a zero speed command, stopping the train. So what the hell happened?

A Type 1 failure is a false positive. A malfunctioning cab signal unit in the train or malfunctioning wayside track detection unit could cause such a thing. The Washington Post mentions a 2005 incident where a Metro train got a clear cab signal into an occupied block, but the operator noticed and was able to hit the emergency brake. No cause could be determined although the Metro does use a type of GRS relay that the FRA issued a warning about in 2000 for being prone to sticking.

If you look at the positions of trains at impact you can get a good idea of where the problem occurred. If the stopped train was well inside the fixed block boundary the probably would lie with the train mounted CSS unit. If the Stopped train was just past a block boundary then the fault was most likely with the wayside equipment. This is because the former produces a 0 code independent of signaling logic and the later is dependent on signal logic.

Interesting. It is being reported that this Washington Metro commuter train was 2 months overdue in its brake maintenance. This sounds like one of the key issues the NTSB will be focusing on.

A Type 1 failure is a false positive. A malfunctioning cab signal unit in the train or malfunctioning wayside track detection unit could cause such a thing. The Washington Post mentions a 2005 incident where a Metro train got a clear cab signal into an occupied block, but the operator noticed and was able to hit the emergency brake. No cause could be determined although the Metro does use a type of GRS relay that the FRA issued a warning about in 2000 for being prone to sticking.

I think the primary cause was an ATP problem of some type. It happened as high speed, there was no wayside switch problem (split switch) or derailment. I guess the operator could have broken the seal on the ATP cutout switch and thrown it, but that is very unlikely. If she didn't, she couldn't have gotten moving that fast unless there was an ATP failure.

Certainly an ATP equipment problem is possible, but posts elsewhere on this Forum state that wayside signal modifications were underway in the area. If that is correct I think that human error associated with those modifications, testing and cutover is at least as likely to be the cause as hardware failure.

Folks, I'm going to confine discussion of this latest unfortunate incident involving WMATA over at the newly renamed "Red Line Collision between Ft Totten and Takoma on (6/22/09)". Thanks