I'm kind of shocked that this is even possible. Don't modern rail systems like Metro have automatic stops designed to prevent collisions? Would those stopping systems need to have failed for something like this to happen?

by jfruh on Jun 23, 2009 7:21 am  

That's one of the going theories. During rush hours, Metro trains operate in an automatic mode, with failsafes designed to avoid this sort of collision. Either the failsafes failed (no pun intended) or the operator overrode them.

by Froggie on Jun 23, 2009 7:26 am  

Most accidents have multiple contributing factors but one of the main questions is why the automatic train protection system didn't work. Whether the train was being operated in manual or auto--the atp system should have prevented this type of crash. In the 1996 fatal accident, there was ice on the tracks so that the train slid far beyond where it should have stopped while yesterday the tracks were dry. I suspect trains will have to be operated in manual (slower, more stops and starts) for some time.

by kreeggo on Jun 23, 2009 7:40 am  

Either the failsafes failed (no pun intended) or the operator overrode them.

by Froggie on Jun 23, 2009 7:26 am

Based on what I am hearing - people are saying (initially) that the failsafes failed and that the operator did not act. Both things happened. Of course, this is just preliminary speculation.

Why would an operator override a failsafe??

by Jazzy on Jun 23, 2009 7:47 am  

My thoughts go out to all the victims and their families. And lets not prejudge the operator who died.

That being said, WMATA is a very customer unfriendly organization and I fear that extends to basic safety issue. I often get the feeling they are running Metro for their own sakes, rather than the public. Catoe SEEMS to be doing a good job, but when things like this happen public trust in those institutions that can't communicate with us disappears forever.

by charlie on Jun 23, 2009 7:59 am  

Catoe just announced all trains will be making stops at the eight car stop markers.

by kreeggo on Jun 23, 2009 8:11 am  

On a different note, Fenty just had a news conference a little bit ago and gave a revised death toll of 7...

by Froggie on Jun 23, 2009 8:20 am  

The trouble with "failsafe" ATCs is that train operators tend to space out while they are in operation. The operator has to be conscious only long enough to open and close the doors. If the ATC fails while the operator is spacing, no one is going to notice the rear of the approaching train. That is why locomotives on real railroads have a sensor the engineer must touch every few minutes or the train brakes are automatically applied. It makes sure he stays alert to watch for track signals. I don't know that Metro trains don't have this feature, but I suspect they don't.

by ksu499 on Jun 23, 2009 8:28 am  

@ksu499: The WMATA trains do not have that feature.

by kreeggo on Jun 23, 2009 8:37 am  

Metro has made some serious missteps in getting information out to the public - which is critical in regaining public confidence.

Catoe's appearance on WTOP Radio at 730am this morning was one of several mini-disasters. The anchor asked him about how such a crash could happen - and referenced remarks by an NTSB Board Member on Good Morning American that - if something like this happens with a fail-safe system in place - the fail-safe system is NOT adequate.

In response, Catoe launched into a rambling two minute dissertation of talking points, beginning with the grief that everyone shares over the loss of life. The anchor had to briefly interrupt him several minutes into his answer to get him to sort of answer the question. (Actually, he never answered it, but tiptoed around it.)

(Note to John Catoe: If you don't answer a reporter's question and that reporter has to interrupt you, you heighten the adversarial relationship. The reporter won't tell you to shut up, but will always have the last word.)

Catoe would be extremely well-served to hire a PR Crisis Manager...a professional who can help Metro navigate its way through what is the greatest crisis of its history. There are several excellent managers in this city, along with dozens of high priced PR firms.

He must tell the public what he knows, but also what he doesn't know...and what must be learned from the investigation. Public trust must be regained. The public has not only a right to know, but an absolute need to know.

I can guarantee you there will be plenty of news stories about Metro's dirty laundry in coming weeks. And people with a grudge against the system for its indifferent customer service will be all too anxious to believe whatever they are told. The stories will range anywhere from safety concerns and distracted operators using cell phones and texting (citing the LA Metrolink crash) to the Dupont station manager running an in-site prostitution business.

In such an environment, John Catoe and his staff must be seen as candid and trustworthy. No more tap dancing. Short, direct, and honest answers. "I don't know," and "the NTSB will be looking into that very issue," are perfectly acceptable. So long as the answers are honest.

And - by the way - was there ever a worse start than Metro's website last night? An hour after the accident, it still spoke of "technical difficulties" on the red line. Folks on the WaPo blog picked up on that one.

The Metro public information staff needs help to get through this one. I'm not blaming them. It's simply too big, and probably beyond their pay grade.

by Mike Silverstein on Jun 23, 2009 8:54 am  

Catoe just announced all trains will be making stops at the eight car stop markers.

Yeah I noticed that on my six car train. I was confused at first because I was so close to the escalator when getting off from the last car. Then I realized they must have changed the rules.

by Joshua on Jun 23, 2009 8:55 am  

Excellent comment by Mike Silverstein.

by Over the River on Jun 23, 2009 9:25 am  

Speaking of the Metro website talking about 'technical difficulties' - this morning the system-wide announcements talked about an "incident" at the Fort Totten station, as if we didn't all know what it was.

I think the thing I am most worried about here is overreaction. How many millions of passenger miles have been traveled on this system with no accidents, no fatalities. This is one incident. It seems to involve a probable combination of more or less random occurrences that together led to the crash. Just like in airplane disasters - mechanical failures and operator error. Probably. It's still too soon to really know, of course. But the point is, while I can see adjustments being made, I hardly see the need to permanently slow trains or to add doors to cars yet to be built. Four, six, eight, 108 doors on each car would not have made any difference in yesterday's crash.

Also, I am fearful this will be used as an excuse not to work toward shorter headways on the system as a whole.

It's a terrible crash, and a terrible day for the Washington, DC metro region. But please let's not add to the tragedy by permanently crippling this absolutely vital part of our urban life.

by Josh on Jun 23, 2009 9:27 am  

Josh: Just to clarify, the doors thing wasn't about safety. More doors let trains load and unload faster, which shortens headways and thus increases capacity. What I was trying to say was, the reaction to this tragedy might affect the planning for future cars, and could help or hurt efforts to get Metro to make other, operational but not safety-related improvements.

by David Alpert on Jun 23, 2009 9:34 am  

I was once caught on a smoke-filled metro train that was stuck under the river between Foggy Bottom and Rosslyn due to a tunnel fire. Judging from the way Metro handled that, it was apparent there was no kind of disaster prepardedness plan (or if there was, nobody knew about it). I certainly hope they reacted to yesterday's tradegy in a more rational manner.

by Juanita de Talmas on Jun 23, 2009 9:43 am  

Thank you for your coverage. I especially appreciate the use of strikethrough text to show how the updates evolved. Thank you very much.

by Eileen on Jun 23, 2009 10:16 am  

I'd like to at least get some meaningful institutional change out of this tragedy. Yes, I want Catoe and others in jail, but I want real reform even more -- not just fixes that address the current safety/service deficits, but long-term institutional changes that address the current democracy deficit of WMATA -- a rider advocacy board with the power to set policy, remove obstinate officials, etc.

A good example of this institutional change _not_ happening is the national economic tragedy. Millions suffering miserably, and a lot more suffering to come. There's been no meaningful reform yet -- just fixes to try to 'get people back to work' -- so we can look forward to the boom/bust cycle happening again, and we - the people who will suffer most - have no real control over it. Let's get meaningful control over WMATA.

by Peter Smith on Jun 23, 2009 10:18 am  

These rules are getting ridiculous. WMATA needs to upgrade their operating infrastructure. It should be easy as pie to know how long your train is to stop at the appropriate point. It should be easy to know where every train is in the network and to communicate between all of them so that collisions cannot happen, period. There is no need to reinvent a wheel here. This is all easily doable.

by NikolasM on Jun 23, 2009 10:36 am  

Veteran operators generally keep busy all the time. They are constantly monitoring the track ahead (who knows when an obstruction, incorrectly set switch, unexpected red signal will appear on the right of way?), checking their speed readouts, making the correct announcements, thinking about any special orders for the day regarding the section of track they are operating on, monitoring communications, insuring they stop at the proper profile, making sure they open door on the correct side with all cars on the platform, etc. Newer operators often do this initially and then begin to relax and enter a false comfort zone.

Inexperience also plays a role when encountering situations you have not encountered before. The odds are that you do not have an accident/incident when encountering an unexpected situation but react and then learn from the experience. However, that is not always the case--sometimes with tragic consequences. This is not a comment on this particular accident--I don't know what happened in this instance--just a general observation.

by kreeggo on Jun 23, 2009 10:58 am  

Wasn't sure how to address this and not sound reactionary, but i'm constantly applying this lesson to automobiles, so why not trains? That is, what is an appropriate top speed for trains, for WMATA?

I read somewhere that WMATA trains hit 59 MPH. OK, so what if that top speed was only 40 MPH -- how much time savings are we talking about? And how much safety can we get out of a reduced top speed? If it saves one life over a ten year period, is it worth it, based on the number of people it _might_ chase off the system (keep in their cars)? I'm open to arguing 'yes' -- but I'd like to see some numbers.

I've read several times now of situations where car people want to increase the speed limit along some stretch of road from, say, 25 to 35 MPH. Of course, us walkers and bikers oppose it, so we calculate how much time car drivers would save by upping the speed limit and making the road much more dangerous and inhospitable for walkers and bikers. The time savings is usually very small (measured in seconds or two-minutes-or-so over a three or four mile stretch of road).

So then we start looking at things like dedicated bus lanes, 'green wave' technology for cars and bikes -- things which _increase_ the performance of a road by reducing the stop-and-go nature of traffic -- and don't set autos loose on walkers and bikers.

Maybe WMATA trains could do the same? Express service, limited stop service, etc.?

Cars, to me, are fundamentally anti-human because they travel at speeds that humans were not meant to travel at. The 'controller' apparatus for cars -- be it technological, social, etc. -- has not caught up with the unruly/wilding nature of car technology -- thus the daily slaughter on our roadways.

JHK has said we should fix regular rail before doing all the high-speed stuff. He has a point.

The 'top speed argument' is worth looking at.

by Peter Smith on Jun 23, 2009 11:33 am  

Officials are finding clues and victims in Mondays train crash in Washington DC. Amazing multimedia posted too. http://pfx.me/va

by Information Desk on Jun 23, 2009 1:18 pm  

Peter Smith -
Train accidents are so rare and so unique that a statistical study of crashes to the effect of certain speed levels on safety is simply not possible. What happened in the DC Metro area train network this week - for the first time in decades - happens every single week in the DC Metro area automobile network. I find it very likely that if we blindly reduce the speeds and thus the ridership(even if ridership is only mildly affected), overall transportation fatalities will increase significantly.

I look forward to the results of an NTSB investigation before anyone is able to make an insightful comment on improving the safety of the rest of the fleet - including replacement of the 1000 series.

by Squalish on Jun 23, 2009 3:20 pm  

Train accidents are so rare and so unique that a statistical study of crashes to the effect of certain speed levels on safety is simply not possible.

straight absurd. studying string theory experimentally? difficult. studying train accidents? not so difficult.

happens every single week in the DC Metro area automobile network

if you want to compare WMATA's death rate to that of DC's automobile death epidemic, be my guest, but i'd suggest that is not aiming very high.

we've determined that speed is a critical factor in all sorts of transportation-related injuries/deaths, yet many still continue to defend speed -- especially high maximum/top speed -- as desirable/necessary. car people. train people. whoever. i don't understand that mentality, especially given that there are so many obvious answers to improving average reliability/speed/performance.

I find it very likely that if we blindly reduce the speeds

keep knockin down those straw men.

by Peter Smith on Jun 23, 2009 3:40 pm  

NTSB has announced that the "mushroom" (the emergency braking system) had been activated by the operator prior to the crash and the brakes on the train had been engaged.

by kreeggo on Jun 23, 2009 7:18 pm  

Peter Smith, there are three issues with speed: momentum at impact, asymmetry of structure, and reflex time. Comparing Metro's speed with an automobile' is apples to oranges, considering the difference of conditions.

1: Your reference to string theory means you probably understand that momentum is also a factor of mass. It's one of the reasons hitting another car with a car is more deadly than hitting one with a bike. A freight train at 15 miles per hour can decimate a truck simply because it weighs 5,000 tons. Should we not also severely reduce the size of buses and trucks to reduce fatalities? The only plus mass has on speed is that mass doesn't affect reflex time - except that in real world-situations, higher mass means slower stopping speed.

2: The way the crashing objects are built also matters considerably. Two cars hitting each other at 35mph is a lot safer for the drivers, who are protected from the incident by devices that reduce both the force and the impulse felt by the driver (such as a crumple cage or airbags). Pedestrians and bicyclists get no such protection, which is why hitting them is so frequently fatal. A railcar might be less safe than a car, but again, speed is only one issue in crashes.

3: Reflexes, a human problem, cause more accidents than speed alone. Human reaction time is 3/4 of a second or somesuch, which must be added to braking time, and can be clouded by drowsiness, darkness, laziness, etc. Some people are just bad drivers and some highways are unsafe for their speed. The number of factors in play even for an automobile collision are an order of magnitude different from a subway collision.

In a street, there are other drivers, pedestrians, cyclists, vesparisti, and segway dorks all over the place. There is just more stuff to hit - including all of the infrastructure. In the air, where planes regularly go over 600 mph, there are so few things around, that a pilot would have to be willfully negligent to hit another plane at cruising altitude. Unsurprisingly, the vast majority of minor airplane incidents happen at airports.

Metro is far more like 30,000 feet than Pennsylvania Ave. there are failsafe measures (not good enough, I suppose), the tracks are designed for speed and the system is designed to limit speed. The drivers are more trained than any private citizen with a DL, and in some places, they aren't even used. The route is separated and secured such that only specific approved vehicles and personnel can enter without assuming liability.

So yes, speed can still introduce more danger, but there are a hundred factors that diminish it. Like any other dangerous thing, you just have to know what you're doing with it. The with 215 trips taken last year, thepassenger-mile death rate of the Metro compares favorably to bathtub electrocutions, heat stroke, and alcohol poisoning. The speed is a risk worth taking, unless you assume that the potential of death, no matter how hard you fight against it, is simply not worth it. That's crazy.

by Sighing on Jun 23, 2009 8:25 pm  

Peter Smith:

Re: your suggestion:

Maybe WMATA trains could do the same? Express service, limited stop service, etc.?

How exactly do you propose doing that on a two-track system? Using the track going in the other direction to pass the local trains?

Also, please note that the system has been straining to meet capacity. Tell me what effect slowing trains down would have on that problem. Please.

by Just wondering on Jun 24, 2009 3:17 am  

How exactly do you propose doing that on a two-track system?

There are myriad ideas already on the table to increase the capacity of not just Metro, but every transit system everywhere. If one is all amped up on top speed, then one will never look at the obvious options -- like express trains (skip some stops--single track--no problem, or add a bump-out track for local trains to wait while express trains pass), like adding doors, like adding cars, like removing seating, like installing tunnels to connect transfer points to ease congestion, like decongestion pricing, like helping employers to shift operating hours to non-peak times.

by Peter Smith on Jun 24, 2009 9:50 am  

To be fair, I mentioned four doors more as a means to speed for entering and exiting to reduce dwell time (and to slightly increase car capacity). I am not a mechanical engineer, but I expect that having another door would reduce the crashworthiness integrity of the metal walls of the car. And not that many subway systems have 4 door cars, at least in North America.

by Richard Layman on Jun 26, 2009 7:10 am  

WRT to Peter Smith's points about increasing capacity, don't forget that at the end of the day there are real constraints based on switching capacity, platform length and width, etc. So with a two track system, skipping some stops only adds a bit of capacity. I'd leave it to Matt Johnson for calculations but some limited stop service would probably only add capacity for a couple of trains/hour, important, but during rush periods, when express service seemingly makes sense, is when you have to stop at every station.

If anything, the crash shows the need for redundancy, the need to add the separated blue line, and maybe the need to consider doublestacking tunnels. And while the Michael S. brown line doesn't make perfect sense from the standpoint of heavy rail financial justification, the need for redundancy adds another dimension to the argument.

http://www.flickr.com/photos/rllayman/2281813917/

by Richard Layman on Jun 26, 2009 7:19 am