Don't permanently slow Metro for minor added safety

Transit

Don't permanently slow Metro for minor added safety

by David Alpert • May 11, 2010 11:28 am

A driver slammed on his car's brakes on the Capital Beltway yesterday after coming too close for comfort to another car in front. The maneuver successfully avoided a collision. In response, AAA called for all drivers to travel no faster than 40 mph indefinitely.

Photo by technotheory.

Of course, this didn't happen, because it's silly. Drivers come a little too close for comfort all the time. Some of the time, they don't stop in time and collide. Occasionally, that causes injury and even death.

Crashes, injuries, and deaths are terrible, and we should try to minimize it with better road design, especially in areas with pedestrians and bicyclists as well as motor vehicles. But nobody is suggesting making all freeways 40 mph, even though it would drastically reduce fatalities, because it would also curtail mobility more than people are willing to accept.

However, this is essentially what ATU Local 689 President Jackie Jeter suggested for Metro. In a May 7th statement following last week's incident where an operator hit the emergency brake to avoid hitting another train, Jeter called for "Train speeds [to] not exceed 40 miles per hour until Metro has resolved the train circuit issues."

The train operators are understandably concerned that their train is going to come too close to another train, which puts them and the passengers in danger. Nobody wants trains to hit. But slowing all trains down considerably and indefinitely isn't the answer.

At the recent streetcar technology forum, transit planner Thomas Hickey pointed out, "The safest train is one that does not move." Hickey argued that systems can't completely eliminate risk, but they can manage risk.

Talking to the Examiner's Kytja Weir, "Jeter said riders probably wouldn't mind if they knew why [slowdowns] were occurring. She said Metro could use its network of electronic signs in stations and trains to alert riders of a safety slowdown on the affected line."

As a rider, I would mind. Many riders would surely switch to driving, simply because Metro would get slower and less convenient. Plus, the slower trains go, the lower the overall capacity of the line, making them more crowded. As Weir notes, trains are already slower, less reliable, and more crowded just because of the manual operation, which Board member Chris Zimmerman has criticized.

Would things be worse if another train collision happened? Maybe, but we shouldn't cripple our transit system for what could become years just in an effort to avoid all risk. Even with the 2009 crash, Metro is still about 34 times safer than driving.

I've worried that FTA oversight could bring speed restrictions that harm transit and overall commuting safety, like FRA regulation did to intercity passenger rail. FTA Administrator Peter Rogoff assured us that this wasn't their plan, though we never know what a future FTA Administrator might do.

The fact is that there was no crash. A train got a little too close, and yes, that's a problem which deserves the highest priority attention. But this is precisely why we have a human, and why Metro says they think it's wise to keep the trains running on manual, to ensure operators are paying closer attention. The backup system is the human, and in this case, the backup system cut in to prevent a crash.

I feel safe enough riding the trains with the signals, where Metro is working hard to eliminate any flaws, plus the human set of eyes from a trained operator at the controls. It's not going to eliminate every possible source of risk, but there's risk in everything, especially movement. Let's keep Metro trains moving so they can get people where they need to go quickly and efficiently as well as safely.

Jeter also called for better procedures for notifying the union and workers as soon as incidents happen and following up 24 hours later. Those procedures make complete sense and Metro should indeed ensure immediate and thorough communication, to the union, safety oversight officials, and the public about any incident.

31 comments

Tags: labor, Metro safety, WMATA

Tri-State Oversight finds safety lapses on track right-of-way (Jan 7, 2010)
Latest Metro crash not such a surprise (Dec 2, 2009)
Can oversight help transit safety without hurting transit? (Nov 16, 2009)
7000 series designs sacrifice capacity for vague safety (Jul 22, 2010)
Metro crash roundup (Jun 23, 2009)

David Alpert is the Founder and Editor-in-Chief of Greater Greater Washington and Greater Greater Education. He worked as a Product Manager for Google for six years and has lived in the Boston, San Francisco, and New York metro areas in addition to Washington, DC. He loves the area which is, in many ways, greater than those others, and wants to see it become even greater.

Comments
Add a comment »

Jeter's proposed speed reduction doesn't solve anything.

If poor signals and communications systems failed, those should be addressed. If the operator did a poor job, employee training should be addressed. A speed reduction would only lower the bar for Metro performance, creating a convenient excuse for avoiding the actual problems ("Rather than fix systemic problems, let's just make it easier to accommodate them!").

This proposal dumps more troubles on riders, hides Metro's responsibility to identify and fix systems, and ignores Jeter's responsibility to make sure that her members are capable enough to safely address reasonable on-the-job demands (like being able to pull a brake).

Let's keep standards high and bring Metro's performance back up to them, rather than let 689 take the easy road and pull them down.

by D on May 11, 2010 11:47 am • link • report

David and I have already had this discussion offline. Its not my area of responsibility or expertise but I would note a few things.

This is the 3rd incident where an apparent failure of the signalling system caused operators in revenue service to brake to try to prevent collisions within the last year. One operator was unable to do so resulting in nine fatalities and many injuries. This does not count the numerous other signal failures that have been identified where random chance did not result in an incident.

In the latest incident, the operator was reportedly able to stop the train within about 115 feet of the train ahead. I don't know how fast the train was going when the operator observed the train ahead. If it was traveling at the maximum permissable speed, then that's about 90 feet per second.

As has been noted in a previous GGW post, you can't rely on line of sight at the speeds that trains are moving. There are too many blind spots.

David doesn't like the solution the union proposed--slow trains until the problem with the track circuits has been identified and corrective action taken. Okay, so the solution is keep doing what we're doing now?

That's exactly what WMATA did in 2005 when 3 trains came perilously close to colliding with each other near Foggy Bottom when the signaling system failed for reasons which were later attributed to a faulty cable. The trailing train stopped within 35 feet of the stopped train. The 3rd train stopped within 12 feet of the trailing train. WMATA said exactly what David is saying now. The system works. The operators stopped the trains. Let's not incovenience people. Four years later this mistake was paid for in blood.

WMATA is working on a system that will provide real-time alerts to signal failures. It may, along with identifying underlying causes of signal failure, ultimately solve the problem. Unfortunately that system is not ready to go today.

So, if you don't like the union's solution, I'll ask you again to come up with another. But that can't be just doing what we've been doing.

by Craig Simpson on May 11, 2010 11:49 am • link • report

Craig: I think the difference is that in 2005 Metro said the system worked, but then didn't go and fix the signals. If they'd started on an aggressive program of rectifying the problems with the signals then, they might have been finished by 2009 and not had the crash. Their error was in saying the system worked, then not taking further measures.

I think Metro should do everything possible to find out why the signal failed, fix it, and prevent this from happening to other signals. They probably need a backup system like BART put in, if the problem here was the same WEE-Z bond problem as at the Fort Totten crash.

I am not okay with Metro simply accepting that this problem might occur. However, I don't think they should slow the trains down until they fix it. Fix it right away, and have operators keep watching out carefully to hopefully avoid any problems while it's being fixed. If there's a specific area that seems to be problematic, slow down there, but not everywhere.

by David Alpert on May 11, 2010 11:58 am • link • report

One reasonable solution would be to use the same number of cars-per-line but make ALL trains 8-car trains.

This has several benefits:

1 - It would allow Metro to maintain the same passenger capacity per hour while INCREASING the distance between trains, thereby reducing the risk of collision

2 - It would allow Metro to maintain the same passenger capacity per hour while DECREASING labor costs. This is because the electricity per car is a fixed cost, but 8 car trains decrease the marginal cost of labor per car. The money saved could be used for things like...fixing signals and communications systems.

3 - Parking trains at the ends of stations won't lead to dangerous situations with passengers running on platforms to catch up with trains, such as at Gallery Place.

This option was never fully explored. Jeter would not agree to this because it would require reducing the number of operators.

As a point of clarification, Craig is a representative for ATU 689, no?

by D on May 11, 2010 12:06 pm • link • report

I don't think that a blanket speed restriction is the way to go, but I also disagree with David A.'s blowing off of the solution proposed by the Union.

There is clearly a problem with the ATC system. Until it is fixed, we can't be sure it will prevent crashes. A speed restriction would increase the response time, but would not necessarily prevent all crashes.

Besides, if the speed restriction causes riders to switch to cars, it is likely that the total number of transportation-related deaths and injuries will increase. This is because driving is a lot less safe.

We need to solve the underlying problem. Until then, treating the symptoms is going to become more and more costly.

I don't think riders will blindly accept a 40mph speed restriction, even if they know what it's for. It will result in more crowded trains, longer commute times, a reduction in mobility, and, ultimately, a loss of riders.

But we can't afford to blindly reject ideas, either.

by Matt Johnson on May 11, 2010 12:09 pm • link • report

I'm always impressed by the contributors to this blog that are able to access and crunch numbers about Metro to produce things like rider statistics for every station and analyze peak-of-peak fare adjustments and headway time changes, etc. But here I'd like to see some data about what a 40mph cap would really mean to train times, not just "everything is risky so why bother changing the speed."

And really, if there was another fatal Metro crash, "maybe" things would be worse? It would be Armageddon for Metro if another controls related fatal crash occurred before the NTSB report and solutions are implemented.

by Lou on May 11, 2010 12:11 pm • link • report

Matt: Thanks. Can you explain the difference between my "blindly rejecting" the solution, and your comments agreeing that it's the wrong solution? I thought I wasn't blindly rejecting but rather rejecting with sight.

by David Alpert on May 11, 2010 12:11 pm • link • report

If we put in a 40mph speed restrictions, perhaps that will light the fire under Metro and the funding jurisdictions to get whatever needs to be fixed, fixed.

At my work, we have safety systems. When those safety systems aren't functioning properly, we are not allowed to operate. Therefore, when the safety systems have problems, everyone works to get them working again.

by Michael Perkins on May 11, 2010 12:13 pm • link • report

@lou: It would probably reduce the average speed of trains from about 27 mph today down to the high 20's or perhaps even lower.

by Michael Perkins on May 11, 2010 12:14 pm • link • report

Sorry, high teens.

by Michael Perkins on May 11, 2010 12:14 pm • link • report

@David Alpert:
I think my reaction was more to your tone, than anything else. Sorry if I came off disrespectfully.

by Matt Johnson on May 11, 2010 12:18 pm • link • report

David: I understand your concerns and there are certainly people who know more about this than I do. I will observe that the signalling problems that I am aware of do not seem to have a particular pattern. There are often crews dispatched that can find no trouble and cannot duplicate the failure. The cause does not always seem to be the same. This is an extremely challenging problem. WMATA now slows trains when a problem is identified but cannot predict where problems will occur or re-occur. While no one wants to reduce train service for any extended period of time, I can't imagine the repercussions of another collision as a result of signal failure. I think all of us are open to other solutions if they can be identified.

by Craig Simpson on May 11, 2010 12:18 pm • link • report

@Michael Perkins

What fire? No action will happen until the NTSB is done with their report at a minimum - because we won't fully know the scope of the fix until then.

by Alex B. on May 11, 2010 12:19 pm • link • report

Dave's analogy is a bit off. It isn't about reducing speed on the beltway; it is about reducing speed when your entire system of red lights go out.

Conversely, making Metrorail more painful to ride isn't going to really change WMATA behavior. It will just encourage riders to move to cars. Having your signals system fail seem to me like about the biggest fail a rail system can deal with, and they haven't done anything with it for years.

Is there a way to identify the segments that are blind spots? Perhaps put a restriction in place for the underground segments? What I worry is speed limits may become permanent.

by charlie on May 11, 2010 12:22 pm • link • report

Railway signaling isn't exactly rocket science. We've been doing it *safely* for over 100 years.

Why is Metro's system failing in an unsafe manner? Signals should be designed to be passively safe -- that is, if something goes wrong, or there is an ambiguity, the light turns "Red," telling the driver/ATC to stop. Signaling failures should result in lots of delays -- NOT lots of crashes. There's clearly a fundamental design flaw in the system that needs to be addressed.

by andrew on May 11, 2010 12:23 pm • link • report

This is the same situation we have with escalators: many years ago there was an accident where someone had an injury; something got trapped by the moving steps. I don't remember the details, but I do know that afterwards Metro decided to slow the speed of all escalators. (They also added the line of brushes along the edge to discourage folks from standing too close to the sides.) You think the descent into the Dupont Circle station is interminable? It used to be slightly faster. For folks who regularly commute via Metro, this slowdown is highly annoying - as one of the few who remember the faster days, I am more annoyed than most.

Next time you're riding Metro to Pentagon City mall, compare the escalator speeds of Metro's sluggish escalators with the ones in the private mall. I think we should be able to match their speed.

Please, Metro, turn up your escalator speed!

by Michael on May 11, 2010 12:24 pm • link • report

@D I do political and legislative work for the union. This subject is outside of my area of work within the union and my comments here reflect my own views.

by Craig Simpson on May 11, 2010 12:24 pm • link • report

If the average speed went from 27 to 18, that's 33% slower, or half again as long to get somewhere. A 20 minute trip would be 30, on average. So that's some perspective.

by Lou on May 11, 2010 12:31 pm • link • report

Matt: No need to apologize. I was just trying to understand your reaction.

by David Alpert on May 11, 2010 12:32 pm • link • report

Isn't it weird that trains are riding on manual because the technology does not work, while we do not trust train operators to remember how long their train is?

by Jasper on May 11, 2010 12:52 pm • link • report

I don't understand why the fact that a driver had to slam on the brakes is reason for slowing speed across the system: this happens tens of thousands of times a day on the roads. The fact that 9 people died about a year ago does not alter the situation: that was the worst accident in decades and the fault was signalling not speed. Most importantly, many more people will die if slow speeds on the Metro force people to drive.

Employee safety is very important, but the most important function of Metro is to deliver a service.

by SJE on May 11, 2010 12:54 pm • link • report

@Lou: again that's a guess rather than a calculation. I know the average travel speed because I finally got my analysis code working this morning, but I don't know what the resulting speed would be for the reduction in maximum speeds.

The 40mph figure seems like a rectal pluck. There should be a better way to figure out what the allowable speeds are given sighting distances. Only problem is that those speeds would probably be intolerably low.

by Michael Perkins on May 11, 2010 12:56 pm • link • report

I'm glad Matt and David are getting along now.

by D on May 11, 2010 12:56 pm • link • report

"If we put in a 40mph speed restrictions, perhaps that will light the fire under Metro and the funding jurisdictions to get whatever needs to be fixed, fixed."

Are you kidding? What fictional world do you live in? This is an agency that killed nine people last year, and has spent the time since then sticking its hands over its ears and screaming "lalalaIcan'thearyou" at the top of its lungs. Show me one thing that they've done after killing nine people that is even remotely scientifically validated to reduce a crash.

Bonus - the captcha is "mission fail". Well said, recaptcha, well said.

by varun on May 11, 2010 1:32 pm • link • report

@varun:

Oo! Pick me!

They've failed to fix elevators, escalators, and signals; they have had more trains break down; they have increased fares. That will reduce ridership to the point where there won't BE a Metro, which will DEFINITELY reduce the risk for a crash.

by D on May 11, 2010 1:41 pm • link • report

I just disagree with thinking that humans are the solution when the problem is the train control mechanism. Upgrade the train control mechanism so you can eliminate these troubling humans once and for all. I would love to not have to pay their weighty salary/pension/overtime/anything and be able to have frequent service all through the night.

by James on May 11, 2010 4:04 pm • link • report

@varun: they have started to test the track circuits more frequently for anomalies.

I win.

by michael perkins on May 11, 2010 6:09 pm • link • report

I get really tired of Union officials being seen as primary representatives for "alternative" positions on issues with regard to the service they work for. This is true for the transit workers union but I absolutely feel the same way about Kristopher Baumann and the DC Metropolitan Police Dept.

They represent their members first and foremost. Often this comes at the expense of the people who are allegedly served by their members.

That doesn't mean I don't appreciate their position and knowledge on issues. But I don't see it as the last word either.

by Richard Layman on May 12, 2010 8:40 am • link • report

But nobody is suggesting making all freeways 40 mph...because it would also curtail mobility more than people are willing to accept.

there are several problems i have with this sentiment.

the first is that is just echos that same old "can't happen here, we're different" blabber that is all too common, and completely nonsensical. and yet, in spite of however many times we hear, "that can never happen here" -- that happens here. i was last proved wrong, even if my boast was more in jest than seriousness, when someone decided to actually stripe bike lanes on Pennsylvania Ave. what people are 'willing to accept' is a complex combination of factors, and it starts with leadership -- someone has to be willing to say the current situation is not acceptable.

what are 20 MPH zones besides something "than can never happen here"?

a separate objection, tho, is a long-running problem i've had with the author's point of view and/or this blog in general -- what seems to me to be a general callousness towards the sanctity of human life. if we decide to treat every life as sacred, we might start to see some results. but that will be difficult to achieve when declare that the sacredness of life is less important than mobility. who knows -- someone might even have the audacity to suggest that we can achieve both safety and mobility at the same time -- ftw! or, someone could get really crazy and suggest something...counterintuitive -- like, removing lanes to improve the flow of traffic. who knows? maybe reducing traffic- and death-causing accidents could do the same?

The leading killer of young people in America today!

No -- we need maximum speed to achieve mobility!

Brilliant!

by Peter Smith on May 13, 2010 2:26 am • link • report

Above, Craig Simpson wrote that "In the latest incident, the operator was reportedly able to stop the train within about 115 feet of the train ahead."

"Reportedly"? Not in Metro's report. Metro says the trains stopped 600 feet apart. http://www.wmata.com/about_metro/news/PressReleaseDetail.cfm?ReleaseID=4450

Am I mixing up two different incidents?

by Turnip on May 16, 2010 1:22 pm • link • report

Just thought this was relevant -- Copenhagen not quite at 0 road deaths, but they're still trying to get there:

http://politiken.dk/newsinenglish/article965600.ece

police oppose a lower speed limit.

and we don't have to _just_ slow down trains -- we slow down _everything_ that is not conducive to a high quality of life, and everything that is not conducive to staying alive -- including cars.

by Peter Smith on May 19, 2010 3:55 am • link • report