Recent accidents have renewed calls for platform edge doors in New York City.
For the second time this month, a gruesome death in the subway occurred when one rider pushed another into the tracks and into the path of an oncoming train. The latest incident happened at 40th St. along the 7 line, and the suspect, with a history of mental problems, said she was targeting Hindus and Muslims as revenge for the 9/11 attacks. After an early December incident was caught on camera, this weekend’s attack was no less disturbing.
After the horror of the incident fades away, the reaction to these deaths focuses around platform doors. In various other cities across the world (but not all of them), platform doors are a common sight. They keep people and debris off the tracks while providing the option to air condition underground platforms. With the MTA focusing on the number of people who were hit by trains — usually around 150 per year — the agency has seemingly brought this response on itself.
But should we be so focused on platform doors? Can we? Those are the two questions that require a rigorous answer before anyone moves forward with the idea. Already, though, forces are building. While some at the MTA have called the doors a non-starter and two capital projects — Second Ave. and the 7 line — have seen them axed from initial plans, a new faction within Transit will at least entertain the idea. “We’ll have to revisit it,” Transit President Tom Prendergast said to The Daily News this weekend. Pete Donohue had a bit more:
The MTA has had on-and-off discussions about placing protective barriers along at least some platform edges, as other cities like London and Paris have done. In 2010, the MTA released a Request for Information for a pilot program, but nothing ever came of it. A New York-based company called Crown Infrastructure submitted a response to install the doors for free, as long as it could collect revenue from LED video advertising on the barriers.
“We haven’t made a conscious decision to table it and not do it at all, but we haven’t made a decision to keep it going either,” Prendergast said. “It’s suspended animation.”
MTA officials also said in 2007 platform doors would be installed on the 7 train extension, and that they were considering doing the same for the new Second Ave. subway. While both projects are under construction, platform doors will not be installed, said an MTA spokesman Friday, calling them “cost-prohibitive.” It would cost an estimated $1.5 million to install sliding doors along two platform edges in a new station, and more to retrofit an existing station. The MTA has 468 station, although many are too narrow for doors.
In an official statement released to Ravi Somaiya of The Times, the MTA said: “Based on the MTA’s preliminary analysis, the challenge of installing platform edge barriers in the New York City subway system would be both expensive and extremely challenging given the varied station designs and differences in door positions among some subway car classes. But in light of recent tragic events, we will consider the options for testing such equipment on a limited basis. Of course, we remind customers of the overall safety of the subway system but urge them to stand well back from the platform edge and remain watchful of their surroundings.”
I’ve written on the aspect of costs before, and the equation remains similar. These accidents happen about once every 12.5 million passengers. Very few are caused by another’s push while many happen due to the carelessness of someone on the platform and others simply as accidents. The cost of the doors vs. the cost of saving a life is a delicate balancing act.
That’s the answer to “should we,” and the answer to “can we” is far more difficult. Outside of the factors mentioned in the MTA’s statement — station design, variable door placement on rolling stock models — there is one overarching problem. Platform edge doors generally benefit from automatic train operations systems to ensure the doors lineup properly. Not every platform door system needs ATO, but those with ATO work better and faster. The MTA is countless years, dollars and union fights away from an ATO solution.
So with politicians and the architect of the JFK AirTrain all angling for platform edge doors, Transit will take a look at the technology. I doubt we’ll see much movement over the next few years though as the organizational and financial challenges are too great to overcome for a systemwide implementation.
76 comments
A random argument and an emotionally disturbed person will now necessitate hundreds of millions of dollars in expense and result in a more claustrophibic/etchittied environment on the platform.
Agreed with this.
I’d add: if you really care about safety, push to reduce automobile usage and use the money to expand subway usage. More people than 150 are killed in surface traffic accidents, and even more injured.
Even the DO IT 4 TEH CHILDRENZ!!11! crowd should agree, but for that they didn’t have their heads so firmly lodged in their asses.
The DO IT 4 TEH CHILDRENZ!!11! crowd is lobbying for rear-view cameras on cars. Because paying attention when backing out of a drive-way is too much work and unAmerican.
Car redesigns which make cars more aerodynamic have actually made rear visibility worse. Trucks have always had awful rear visibility. SUVs… the same.
So at this point I think the rear-view cameras are becoming necessary. Well, thanks to the Charge Coupled Device, they’re cheap now.
I’m for the cameras. I just find the faux outrage behind the motivation to be hilarious.
OK, agreed with that.
Some Platform-Edge-Doors go floor-to-ceiling, allowing for true climate control in the platform-space. Others do not, providing a safety/debris barrier but not the ability to condition (the latter also avoids the issue of high-pressure changes when trains arrive and depart stations).
My concern, which I never thought of until recent events, is about drainage. Right now, most station drainage happens via the tracks. Water (hopefully from mops) on the platform makes its way to the track, where it gets channeled to a low-point and is pumped out. What happens if that flow of water is stopped by floor-to-ceiling PEDs, like those in the above photo? In an event like Sandy, it could make the station look like a fishtank while the tracks remain relatively un-flooded. Maybe that would be a good thing; but I’m not quite sure.
But adding adequate floor drains to an old station platform is quite a daunting task.
The barriers would never be designed to withstand the hydrostatic pressures of filling up. They’d need ridiculous bracing – or preposterous thickness – to not blow out – simply increasing the cost of tidal catastrophes like Sandy.
No doubt the bottom 6 inches or so would simply be open – which would mean that debris would still find it’s way to the tracks – but perhaps less mobile phones – and less bodies in need of medical supervision.
Sounds like a perfect pretext for pissing away a few billion dollars in studies and pilot programs – before concluding what we already know. There are too many differences between rolling stock, divisions, systems and platforms to ever make it as practical as simply promoting better practices of safety and awareness. And we all know that those billions can be better spent.
And for those bent on suicide, they would just find another way anyway – so no lives would be saved per se, only saved from delaying trains in the process.
If CBTC is ever installed and implemented on the B division lines, it can cause even more delays. So what’s your point?
If English is your first language – and you read my post – then my point would be evident.
There are better ways to spend billions of dollars on our subway system then even contemplating platform doors.
Such as?
More trains to get crowding off the platform, make transit more reliable, and move more people.
Standardizing the B division rolling stock design to a standard door layout — and making it WALK-THROUGH like the London Underground’s new trains — would be a fine thing to do. It’ll take 20 years. And it’s a necessary prerequisite for any sane implementation of platform screen doors, so I suggest directing all the “platform screen door” money to that. 😛
They seem to function under the assumption that a mix of 75′ and 60′ equipment is worth it if using the former where you can and the latter where necessary. Well, that’s my take from reading comments in a few different forums anyway. (Someone here commented that they plan to buy more 75′ cars, in which case it might take closer to 40 years.)
The A division, however, might be easier to standardize, and does include some of the more crowding-prone stations.
The R179 order is 60 foot cars, so there are currently no plans to buy new 75 foot cars.
Since the design draft of the R211 order is not complete, this just necessitates that they use 60 foot cars for it as well. Even if it takes years or decades, yes, transit should install platform screen doors.
Its not as sexy as building new train lines here and there, but why bother to do that when you can’t upgrade what you possibly have?
In fact, if platform screen doors were in place, why not upgrade the systems signals so the trains can run faster, close to the speeds of the LIRR or the Metro North?
Since the 1970s, the only 60-footers that have ever been built are those that are required for service on the BMT Eastern Division (R143, R160, and R179.) I’m not sure that the R211 will be 60 feet long. Hopefully, the MTA will stick with 60-footers, but it’s not guaranteed.
But if the R211 order is for cars 60 feet long, then faster service and stationary PSDs might be an option.
If you can just upgrade the signals so trains can run faster without a meaningful tradeoff in safety elsewhere, do it. Increasing train throughput on a very busy line can be the equivalent of adding a few trains per hour, which in turn reduces platform crowding and improves platform safety.
Expansion of the system.
Although new subway stations can be designed for full-height screens and doors, other cities around the world have found it much easier to use half-height screens when retrofitting older stations. 5ft to 6ft seems to be fairly typical. One issue can be the need to strengthen the platform edge to carry the weight, particularly if the edges were built with a slight overhang.
A very good example is Paris Line 1, which is the city’s oldest and busiest line, with lots of tight curves, awkward platform layouts and so on. This has been successfully fitted with half-height screens over the past 3-4 years to permit a switch to automatic operation. ATO is not a requirement – the doors can work successfully with manual driving. Line 1 currently has a mix of automatic and manual trains inter-working quite successfully. London’s Jubilee Line Extension also has used full-height screen doors with manual driving for more than a decade.
As well as preventing accidental and deliberate falls onto the track, platform doors help to improve dwell times, enforcing more disciplined boarding and alighting, with benefits for service reliability.
One complication is the need to ensure a proper alignment between train and platform doors, which can restrict the operator to one type of rolling stock. It’s less of a problem on stand-alone lines, but maybe a challenge for a large system with different car types. Although Helsinki has a relatively simple line, its automation project ran into problems because the two main car types have a different door spacing; this was too much to be resolved by using slightly wider screen doors, a solution adopted elsewhere.
It’s worth noting that even Paris’ manned trains are automatically piloted, the “drivers” simply act as conductors.
I wouldn’t be surprised to see the MTA give in and test platform doors on some IRT station or on the L if the reports of people either falling, jumping or being pushed onto the tracks become a high-focus issue with the local media due to the 49th Street and the 40th-Rawson incidents. They’ll do it if for no other reason than to be able to show the media and the public they’re doing something, though any doors will have to be at least as reliable as the Union Square platform extenders, and the MTA would still have to explain to the B division riders why none of their midtown stops can get doors until the final 75-foot cars are retired from service.
Considering vehicle construction and MTA finances, R46 will run until 2020 and R68 will roll on into the late 2030’s. That will be quite a long wait.
They could do a partial solution. Add wrought iron or galvanized steel fencing in between the yellow edge and the orange caution stripe. Have spaces (6′-8′ wide) in the fence for doorways. Having the rest of the platform edge can cut down on deaths. No moving parts to maintain or glass to repair/replace. Use bars for elevated steel structure platforms to cut down on wind loads. The other platforms can see panels with advertisement or system information.
P.S. Full height platform screen doors need platform CCTV. Otherwise, you can have…unfortunate accidents.
No, it’s perfectly possible to use PEDs with both 75-footers and 60-footers (see my comment on movable barriers below.)
But remember, we’re talking about an agency that can take months to fix an elevator or escalator. Whatever is put in would have to rely on the most basic possible technology to work, like the moving platforms at Union Square. Start complicating thing to the point where the platform doors have to know if an R-46 or an R-160 is entering the station so it can adjust it’s locations where the gates would open, and you’re just asking for problems (and, for the MTA, lawsuits, come the day when a platform door for a 40-door, 60-foot train opens when a 32-door, 75-foot train is in the station, and some poor schlub walks into the opening and then down between two railcars).
Not every platform door system needs ATO, but those with ATO work better and faster.
I can’t think of any line with platform doors that doesn’t have ATO. Can you? Chris J mentions that some trains on Paris’s Line 1 and London’s Jubilee line are “manual,” but I believe he’s confusing ATO (where breaking and acceleration is done automatically, but doors are opened manually) with full-on driverless (where there is no human on board, or at least no human on board who’s controlling the train).
I also have a feeling that the MTA is not being totally truthful about their reasons for not installing platform doors on the new 7 and SAS stations. They say it’s about cost, and then cite a figure of $1.5 million per new station, but this leaves out ATO entirely. In reality I think the main barriers are signaling costs and, frankly, their own incompetence at implementing it – neither of which they’re willing to own up to in the press.
This is why I said on Twitter that I think Prendergast is either lying about them “considering” it, or he’s ignorant about the amount of automation that would be necessary to implement platform doors and how far behind the MTA really is from doing it. For the sake of the subway I hope he’s lying, but my fear is that the MTA is so disconnected from reality that they don’t even realize how far they are from implementing platform doors. Evidence for the second possibility is the “request for information” they put out for platform doors (which, naturally, didn’t get any bids). If they actually knew that it would involve ATO, why would they put out a request like that?
So there’s a third possibility when it comes to Prendergast: He’s playing politics/PR and is saying what he knows people need to hear in order to avoid a City Council freakout.
As to the Request for Information returning 0 bids, well, that’s what an RFI is supposed to do. It’s an information-gathering tool used to ascertain whether an idea is technically feasible, and in this instance, it did return a reply from a company that proposed capturing all ad revenue on the doors in exchange for some of the installation rights. The MTA hasn’t moved to the RFP process in which bids would actually be solicited.
So there’s a third possibility when it comes to Prendergast: He’s playing politics/PR and is saying what he knows people need to hear in order to avoid a City Council freakout.
How is this distinct from “lying”?
They’ll spend 20 minutes “studying” it and then issue a report that it costs too much and isn’t practical. It’s not lying. It’s just not taking these cries seriously.
Citing a cost of $1.5 million per new station without mentioning the need for automation (unless I’m wrong and platform doors don’t necessitate ATO these days? but nobody has presented me with a counterexample yet…) is pretty damn close to lying. It’s a lie by omission, at best.
Is there some reason that some trains or stations have more accurate stops than others? They always seem to stop in the same place to me… in some stations the motormen press a button on a console outside of the train far smaller than the platform doors on the Airtrain.
Im just nitpicking about the need for ATO to get platform edge doors, do not mistake this post for supporting the doors.
Then you have situations like the Sixth Avenue platforms at 34th St, which at random points in time host ten-door and eight-door trains. The platforms are much longer than they need to be as well, so overshooting is okay (sort of).
With PSDs this leeway provided by the platform would be gone.
Again, movable PEDs would be really useful.
https://d13uygpm1enfng.cloudfront.net/article-imgs/en/2011/10/20/AJ2011101915245/AJ2011101915246M.jpg
AFAICT, that just means that they can be retrofitted if the MTA changes the allocation of cars to a line. It doesn’t mean that you can actually run 75-footers and 60-footers along the same line through the same station with platform doors.
It’s proprietary technology (for now) and I can’t find evidence of an existing installation anywhere in the world, so this should be taken with a grain of salt.
For now, it doesn’t exist. But this is the technology that Tokyo Metro is planning to use on its Hibiya Line, and other lines with through commuter rail services where it’s currently impossible to install PEDs.
Re incompetency: well, hey, if they aren’t doing something because they know they’re incompetent at it, it’s a huge improvement over doing something despite being incompetent at it.
Good point. If Bush had known he was incompetent at invading Iraq…. well, anyway.
The (vast) majority of Shanghai’s stations have full height screen doors, and none of their lines have ATO of CBTC (with the exception of line 10)
Not so. I am well aware of the difference between ATO and driverless. The Jubilee line has had ATO for about 2 years, but ran with conventional driving and doors for a decade before that. Paris has been introducing attended ATO on other lines. All I meant was that it ‘can’ be done fitting barriers without ATO, even if that would be preferable.
Whether there is a budget or business case is another matter, of course. But it is technically feasible.
There is a line in St Petersburg with platform doors with full manual driving. The doors are big and thick and are there for flood control. They’re also not really wider than the car doors so require precise stopping by the driver to work.
AFAIK, none of the Tokyo lines with platform barriers (e.g. Meguro, Nanboku, Fukutoshin) use ATO. They do away with the conductors that are used on most other lines in the region though (leaving only a driver).
ATO very clearly isn’t necessary anyway: Japanese rail practice has always used very precise stops, without any automation (I seem to recall hearing that drivers are disciplined if they stop more than 0.5m out of alignment), because people line up at the door locations marked on the platforms. It’s simply a matter of training.
The Maronouchi line uses ATO and platform edge doors. The Namboku, Chiyoda, and Fukutoshin lines still use ATP with platform edge doors. If it’s possible to use ATP in conjunction with platform screen doors in Japan, then maybe the MTA can do it here. Even with the trains being different lengths, I’m sure the MTA could manufacture movable barriers like the ones they plan to use in Japan.
By the way, the Paris Metro line 13 uses PSDs and is still manually driven.
Nooo, all of Paris’ lines have ATP
I haven’t been in Paris in a while, so I might have been wrong about line 13. The 3bis and 7bis lines still count, by the way.
There’s still a difference. The 1 and 14 lines are completely automated with absolutely no driver whatsoever. The 3 line has CBTC but still has an onboard conductor. All the other lines have a driver who pushes a button to start the trains; even so, the automatic features of these trains are limited to speed control in the tunnels. You are right though, those lines all have ATP.
The MTA also can’t tell the difference between platform edge doors and platform screen doors. Whereas the former is half-height, the latter is full-height.
Knowing the MTA, the screen doors would probably not hold up to the abuse and poor maintenance, so people would either get stuck on trains or the doors would be pried open. Both outcomes would defeat the purpose.
Either that or the doors get vandalised.
Platform edge doors are fine enough. Screen doors are too costly, and besides, most deaths do not occur due to someone trying to get over a 6-foot tall gate.
No one else seemed to notice that in many stations it’d be physically impossible to fit PSDs in. Union Square is too curvy, Herald Square (6th Ave) isn’t flat and so on…
Many more just seem too decrepit.
Most rehabbed platforms these days have raised middle sections to line up platforms with door heights (and because the MTA can’t be bothered to make all of a station wheelchair-friendly).
Retrofitting these would basically force the MTA to spend money on platforms it just fixed, and installing PSDs or PEDs on platforms would basically prevent other platforms from becoming wheelchair-friendly unless they ripped out the doors, which would be another giant waste of money.
The main reason for only raising part of the platform is to avoid having to change the heights at which the existing elevators and escalators attach to the platform. That’s London’s reasoning, anyway.
In London, they have a very specific scheme where the “platform hump” will always contact the same car; if you get on at the “hump” you can always get off at the “hump”. Their next project is to have roll-through trains, so that people who got on at full-length level-boarding stations can get off at the hump even if they were in the wrong car….
Hey, they’re trying. I don’t think the MTA is trying, because apparently their humps don’t always line up with the same car and they aren’t telling anyone which car they line up with. Yeah.
The hump is supposed to always be at the conductor’s position (the middle of the train, typically). This is explained on the website. Are you aware of any locations where the hump is in the wrong place?
If you raised the whole platform, why wouldnt you just have a small hump or ramp to the elevator? The difference is only 6 inches or so.
Ive always assumed the actual reason is that its much cheaper than raising the whole platform.
Oh, it’ll be physically possible (see Paris’s line 14). It will just be mind-blowingly expensive.
No one else mentioned that PSDs have been successfully installed in other curved stations worldwide. It just costs $1 million or so more to install.
When will those in the US learn that there is no absolutely 100% risk-free environment, anywhere? Risk of injury and death is reduced when people attend to their surroundings, heed verbal and written warnings, and stop looking at their damned phones 24/7.
Given the way guns are sold like candy and automobiles get most of the deference on the streets, I don’t think there is much concern for a “risk-free” environment in the USA.
The concern is a liability-free environment. A transit agency has oodles of liability when a train or bus hits somebody, despite all the careful selection and training of employees and the best efforts to make a safe environment for riders. A privately occupied vehicle operated by a driver who passed the bare minimum of training to receive a state license has little liability, and what liability he does have is mostly covered by insurance we all have to pay into. And, as long as there is no drugs or alcohol involved, there probably will never be criminal liability for irresponsible driving.
Not to mention the enormous degree of service interruption when they’ve got to shut down the line to pick up the body parts and “investigate.”
Anybody have a sense of the monetary cost of a “human-related accident” for a busy subway/rail-line?
[From such a point of view, of course, it would make most sense to concentrate any protection efforts on the busiest core lines/areas…]
Speaking of which, you also can’t stop crazy people from pushing other people onto the tracks.
Barriers would effectively do just that. The quest is, is the cost worth it?
If it’s $1.5 million per platform face, then we’re taking about ~$1.5 billion systemwide. At 4% rate of return that’s $60 million per year; at 2% it’s $30 million per year. The insurance value of human life today is ~$5 million, but it grows over time as per capita incomes grow. If the number of annual deaths prevented is more than 12, it’s a good investment independently of all else, and even if it’s between 6 and 12 it’s good at the interest rate for risk-free long-term investment.
Yeah, okay, fair enough. My nitpick would be the MTA probably doesn’t pay the insurance value of human life. It probably pays settlements that are a fraction of $5M, if it’s held accountable each time at all.* AFAIK, liability is often split between the TA and the victim. From your perspective, the barriers would be a generous backdoor subsidy to insurance companies.
I was thinking more of opportunity cost. $1.5B the TA would have to borrow, probably at 4-6% right now, could pay for:
• about a km of SAS?
• the second station on the 7 extension and the entire Rockaway Branch reactivation?
• ~700 60′ cars (which would be needed for the barriers to be standardized, I’d guess)?
I don’t remember how much signaling upgrades are costing, but I’m going to guess $1.5B is a big dent on the A division’s CBTC needs.
* Legislative solution: make sure those who end up on tracks are responsible for their own injuries, assuming they weren’t pushed deliberately
The legislative solution is not a solution at all. Sadly, it’s the same non-solution used for pedestrians hit by drivers in the US, since drivers are wantonly under-insured, and on top of that the police views all collisions as no-fault, unless the pedestrian failed to look both ways for the correct number of milliseconds, in which case it’s the pedestrian’s fault.
It’s not a good solution. It just means we don’t pay higher fares because of other people’s fuckups. Well, if you want to deal with it for real, either the TA needs to be subsidized to deal with it, or it needs to be part of routine health insurance. But do you see that happening?
This should be considered a security threat in that unauthorized personnel are able to get onto the tracks/tunnels/critical infrastructure. Homeland Secutity Federal monies.
I’m all for this…If somebody can find the money to replace the entire B-divison rolling stock with R160-like-door-dimensions subways cars and R-142-a-likes for the A-Division. And rapid installation of CBTC system-wide. And of course the platform doors themselves. And the extra maintenance crews. And the ADA additions that by law have to come if you’re doing a station renovation. So if the people demanding this can find that trillion or so dollars, let’s do it. To be fair, we did blow a few trillion on Iraq, so let’s just tell everyone Saddam or OBL or Iran is behind the suicides or something and Congress will sign off on the funding in a heartbeat.
“If somebody can find the money to replace the entire B-divison rolling stock with R160-like-door-dimensions subways cars and R-142-a-likes for the A-Division.”
Which might not happen until 2027. The R62 and R68 classes will be retired by then.
“And rapid installation of CBTC system-wide.”
You mean, within 15 years?
As I’m joking in that post, I mean to begin immediately and be finished within a few years.
Railfans should be totally against platform doors as platform doors totally ruin any chance of getting decent photos of trains in stations.
Maybe, but they should be totally for driverless trains as they allow very good railfan window pics.
You can probably still get them on elevated sections – a lot of elevated and open-cut stations are probably too narrow to install doors on.
As Henry says above, screen doors aren’t going to be installed in all 469 stations of the NYC subway because of width issues. Besides, there are lots of places on elevated sections and bridges across open cuts where you can still get decent photos.
[…] a renewed focus on platform screen doors following a second subway shoving death in December, that instability has taken center stage in the […]
[…] 4. Kabak, Benjamin. “‘A Screen Door on a Submarine…’” Second Ave. Sagas, December 31, 2012. http://bkabak.wpengine.com/201.....submarine/. […]