6 ms·

20 years ago, the MTRC in Hong Kong started to install platform screen doors to seal off the platforms and reduce their air conditioning bill as most of the coo

by Renaud 8y ago

20 years ago, the MTRC in Hong Kong started to install platform screen doors to seal off the platforms and reduce their air conditioning bill as most of the cool air was of course just going into the tunnels. That investment paid for itself within a few years. Evacuating the heat from the train is another issue that must be addressed but it’s far from being an unaddressable one and there are already many examples of other cities implementing workable solutions. Having doors installed along the platform has a number of advantages beside air conditioning: increased safety (no more suicides and accidents), more platform capacity since people can get right to the edge of the platform, potential increase in network capacity since trains can enter the station faster as they don’t have to slow down early to enter the platform at low speed to avoid passengers accidents. Older networks like London, Paris and New York have a much harder time implementing these measures as they are often very deep and their infrastructure are old and quite vast, but they each have projects to tackle these issues, although in the case of New York, the cost will probably slow down adoption (not least because of the crazy Union requirements).

bobthepanda 8y ago
The problem with doors in New York in particular is that not all generations of rolling stock have doors in the same location. Unlike MTRC, where one line has its own rolling stock and there is no interoperation between different lines, many services, each with different door positions, can utilize the same platforms. And basically the only way to have a door configuration compatible with all rolling stock is to not install any doors at all. You could replace the rolling stock, but all of it is in good operating condition, and the last of the oldest cars with non-compatible doors would be retired in the 2030s. Air conditioning is also a huge challenge; traditionally, New York's subway has been cooled using the piston effect of trains, with frequent grates open to the surface to draw in air. With air conditioning those large, frequent openings are now a curse.
- IIAOPSW 8y ago
  What fraction of the fleet are incompatible? Just shove all the old trains on a single line for now and install the platform doors on that line in the 2030s.
  - bobthepanda 8y ago
    The separate fleets are assigned to individual lines (which are better thought of as services). But there are 36 of them, pretty much all of which interweave with each other. The R, for example, at some point shares track with the N, the W, and the M, which in turn share track with the Q, the D, and the F, which in turn share track with the E, the B, the G, and so on and so forth. And this is before we consider the fact that in New York, trains are often rerouted due to service changes onto tracks they don't normally serve, and sometimes are changed after years of change in ridership patterns; installing platform doors incompatible with certain trains would reduce operational flexibility. Untangling the lines is not possible, partially because the system was designed to provide as many one seat rides as possible, and partially because the current subway is a patchwork mashup of three separate, competing subway systems; if you were to separate the lines, some trips would just not be possible via transfers, and most transfer complexes would not be able to handle the amount of passengers changing trains. To answer how many cars are not compatible, that would be 2025 subway cars out of a total of 6148 (at least according to Wikipedia.) Of these, 625 were built in the late '80s and are not due for retirement til the 2030s. To put that in perspective, the late '80s fleet is bigger than the entire Boston T rapid transit fleet (Green and Silver Lines not withstanding)
    - IIAOPSW 8y ago
      I know about interlining and trunk lines etc. So about a third of the fleet is incompatible right now. I assume the problem is equally bad in both A division and B division (fat trains and thin trains for those reading along). The track sharing doesn't seem like a huge deal in my proposal. Services usually don't share track across trunk lines. I can't think of any platforms that are shared by trains on different trunks. As an example you could probably get away with enforcing A/C/E are new cars only. Worst case you delay installing the barriers on the shared track segments. There's also the obvious solution of using the L and G as the system test case (yet again). MTA should take the opportunity while the L is shut down anyway for the Canarsie tunnel repair clusterfudge. You're right that it would reduce operational flexibility a bit. In theory service reroutes should be uncommon enough for this to not matter. But you and I know better. If I were to draft this as a serious proposal I'd look at reroute probabilities, come up with some way to estimate the cost incurred by inflexibility, and optimize the installation of platform walls to balance this cost constraint.
- jrockway 8y ago
  The IRT and BMT Eastern Division all use cars that have doors at the same place, so that's not an excuse to not try platform doors. You could make the test station one of the L train stations that's going to be closed for two years and install them at a super leisurely pace. (The L is basically completely isolated from the rest of the system, and already requires special rolling stock for CBTC. And it's going to be closed for two years for other repairs.) Even if varying door locations were the problem... Tokyo has many lines that share rolling stock among 3 or even 4 separate rail companies, and they still have platform doors. There are just a lot more doors in the station than on each train. There simply is no interest or money for platform doors in NYC. Even though people fall onto the tracks and are killed, even though many stations suffer from capacity problems because there aren't enough safe places to stand, even though people holding train doors delays every train multiple minutes a day... we're simply allergic to the concept. I've never even heard it seriously proposed.
  - bobthepanda 8y ago
    The major IRT lines save for the 7 operate mixed fleets of NTT and R62/A equipment, none of which share door positions IIRC. The Eastern Division still uses old rolling stock like R42s from time to time. The last time they tried getting rid of cars prematurely, the R179 order came in late and they were forced to run what cars remained had more aggressively, degrading reliability. > You could make the test station one of the L train stations that's going to be closed for two years and install them at a super leisurely pace. (The L is basically completely isolated from the rest of the system, and already requires special rolling stock for CBTC. And it's going to be closed for two years for other repairs.) This actually was a proposal. It was recently killed in favor of bringing elevators to Union Square, a major subway complex that is not ADA accessible. Which is really why they haven't caught on as a serious engineering proposal; the list of things that would be much better value for money is long, such as systemwide ADA accessibility, CBTC signalling, replacement of rolling stock, refurbishment of dilapidated, water-damaged stations, etc. You could easily come up with $50B worth of upgrades, extensions, etc. before platform doors became a priority. And they're not even that feasible. > There are just a lot more doors in the station than on each train. Got a source for this? This is intriguing because I've never heard of something like this. But this also introduces the complexity that the right doors have to open at the right time, lest a blind person walk into a gap between trains and die. Which is a lot to ask from an agency that can barely maintain something as mundane as elevators and escalators.
- danieltillett 8y ago
  How about just using large strip curtains? Hang them from the ceiling and people can just push through them where the carriage doors stop. Should be cheap to install too.
  - bobthepanda 8y ago
    That would neither really restrict airflow to the point where AC would be feasible, or prevent passengers from falling onto the tracks. Not sure what that would solve as a problem.
    - danieltillett 8y ago
      The large heavy strip curtains would restrict airflow, especially ones that are attached top and bottom. Yes they wouldn't stop people falling on the tracks, but a least they could be installed without costing billions.
- Terr_ 8y ago
  One option is to increase the distance between the train and the doors, so that there's enough of a "vestibule" for people to move around in when a mismatch occurs.
  - mislankanova 8y ago
    I have seen platform doors arranged like what you described at an inter-city railway station. Unfortunately this is unlikely to work for hot and gross subway systems like the NYC subway, where platforms tend to be narrow and overcrowded.
- Gravityloss 8y ago
  Every time there's some cool thing shown somewhere in the world, you have people telling excuses why it can't be done where they live. Close to the NIH synfrome but not quite the same. There are some valid concerns, but I often feel the first reaction tends to be quite negative. Maybe it doesn't make a comment people feel would be worth writing to say "great, let's do that here too!"
- c22 8y ago
  Why not just have the doors a few feet away from the train?
  - pintxo 8y ago
    People will end up in this no-mans-land with the train gone, and the doors closed?
  - bobthepanda 8y ago
    In many places there isn't "a few feet" available from the platform.
amarshall 8y ago
The biggest problem is indeed cost. Installing platform doors on the platforms throughout the system has been estimated in the past at US$1 million per station (US$400–500 million) in 2012 [1] and US$1 billion in 2013 [2]. But those estimates seem incredibly low now given more recent per-station costs elsewhere (CA$10 million, €2.6 million) [3], and the fact that NYC subway construction is higher than most other major cities [4]. As for depth, most NYC subway stations are directly below the street. I’d be surprised if NYC builds platform doors before Communications-Based Train Control (CBTC). [1] https://www.wnyc.org/story/283792-nyc-mta-tepidly-explores-platform-barriers-after-subway-track-deaths-again/ https://www.wnyc.org/story/283792-nyc-mta-tepidly-explores-p... [2] https://www.nytimes.com/2013/01/29/nyregion/mta-says-subway-platform-gates-could-cost-1-billion.html https://www.nytimes.com/2013/01/29/nyregion/mta-says-subway-... [3] https://www.thetransportpolitic.com/2017/09/26/the-case-of-the-missing-platform-doors/ https://www.thetransportpolitic.com/2017/09/26/the-case-of-t... [4] https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-construction-costs.html https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...
- scarejunba 8y ago
  But that's probably because NYC pays some multiple of what anyone else would pay. https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-construction-costs.html https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...
  - amarshall 8y ago
    Yes, and I mentioned that (and the same article) in my post :). But even if you ignore NYC being more expensive, the more recent estimated per-station prices in Paris and Montreal would come to around US$1.5–3 billion for all of NYC’s stations. If NYC remains around 6x the price in Paris, then that’s around US$9 billion for platform doors in NYC.

by Renaud 8y ago

bobthepanda 8y ago

The problem with doors in New York in particular is that not all generations of rolling stock have doors in the same location. Unlike MTRC, where one line has its own rolling stock and there is no interoperation between different lines, many services, each with different door positions, can utilize the same platforms. And basically the only way to have a door configuration compatible with all rolling stock is to not install any doors at all. You could replace the rolling stock, but all of it is in good operating condition, and the last of the oldest cars with non-compatible doors would be retired in the 2030s. Air conditioning is also a huge challenge; traditionally, New York's subway has been cooled using the piston effect of trains, with frequent grates open to the surface to draw in air. With air conditioning those large, frequent openings are now a curse.

IIAOPSW 8y ago
What fraction of the fleet are incompatible? Just shove all the old trains on a single line for now and install the platform doors on that line in the 2030s.
- bobthepanda 8y ago
  The separate fleets are assigned to individual lines (which are better thought of as services). But there are 36 of them, pretty much all of which interweave with each other. The R, for example, at some point shares track with the N, the W, and the M, which in turn share track with the Q, the D, and the F, which in turn share track with the E, the B, the G, and so on and so forth. And this is before we consider the fact that in New York, trains are often rerouted due to service changes onto tracks they don't normally serve, and sometimes are changed after years of change in ridership patterns; installing platform doors incompatible with certain trains would reduce operational flexibility. Untangling the lines is not possible, partially because the system was designed to provide as many one seat rides as possible, and partially because the current subway is a patchwork mashup of three separate, competing subway systems; if you were to separate the lines, some trips would just not be possible via transfers, and most transfer complexes would not be able to handle the amount of passengers changing trains. To answer how many cars are not compatible, that would be 2025 subway cars out of a total of 6148 (at least according to Wikipedia.) Of these, 625 were built in the late '80s and are not due for retirement til the 2030s. To put that in perspective, the late '80s fleet is bigger than the entire Boston T rapid transit fleet (Green and Silver Lines not withstanding)
  - IIAOPSW 8y ago
    I know about interlining and trunk lines etc. So about a third of the fleet is incompatible right now. I assume the problem is equally bad in both A division and B division (fat trains and thin trains for those reading along). The track sharing doesn't seem like a huge deal in my proposal. Services usually don't share track across trunk lines. I can't think of any platforms that are shared by trains on different trunks. As an example you could probably get away with enforcing A/C/E are new cars only. Worst case you delay installing the barriers on the shared track segments. There's also the obvious solution of using the L and G as the system test case (yet again). MTA should take the opportunity while the L is shut down anyway for the Canarsie tunnel repair clusterfudge. You're right that it would reduce operational flexibility a bit. In theory service reroutes should be uncommon enough for this to not matter. But you and I know better. If I were to draft this as a serious proposal I'd look at reroute probabilities, come up with some way to estimate the cost incurred by inflexibility, and optimize the installation of platform walls to balance this cost constraint.
jrockway 8y ago
The IRT and BMT Eastern Division all use cars that have doors at the same place, so that's not an excuse to not try platform doors. You could make the test station one of the L train stations that's going to be closed for two years and install them at a super leisurely pace. (The L is basically completely isolated from the rest of the system, and already requires special rolling stock for CBTC. And it's going to be closed for two years for other repairs.) Even if varying door locations were the problem... Tokyo has many lines that share rolling stock among 3 or even 4 separate rail companies, and they still have platform doors. There are just a lot more doors in the station than on each train. There simply is no interest or money for platform doors in NYC. Even though people fall onto the tracks and are killed, even though many stations suffer from capacity problems because there aren't enough safe places to stand, even though people holding train doors delays every train multiple minutes a day... we're simply allergic to the concept. I've never even heard it seriously proposed.
- bobthepanda 8y ago
  The major IRT lines save for the 7 operate mixed fleets of NTT and R62/A equipment, none of which share door positions IIRC. The Eastern Division still uses old rolling stock like R42s from time to time. The last time they tried getting rid of cars prematurely, the R179 order came in late and they were forced to run what cars remained had more aggressively, degrading reliability. > You could make the test station one of the L train stations that's going to be closed for two years and install them at a super leisurely pace. (The L is basically completely isolated from the rest of the system, and already requires special rolling stock for CBTC. And it's going to be closed for two years for other repairs.) This actually was a proposal. It was recently killed in favor of bringing elevators to Union Square, a major subway complex that is not ADA accessible. Which is really why they haven't caught on as a serious engineering proposal; the list of things that would be much better value for money is long, such as systemwide ADA accessibility, CBTC signalling, replacement of rolling stock, refurbishment of dilapidated, water-damaged stations, etc. You could easily come up with $50B worth of upgrades, extensions, etc. before platform doors became a priority. And they're not even that feasible. > There are just a lot more doors in the station than on each train. Got a source for this? This is intriguing because I've never heard of something like this. But this also introduces the complexity that the right doors have to open at the right time, lest a blind person walk into a gap between trains and die. Which is a lot to ask from an agency that can barely maintain something as mundane as elevators and escalators.
danieltillett 8y ago
How about just using large strip curtains? Hang them from the ceiling and people can just push through them where the carriage doors stop. Should be cheap to install too.
- bobthepanda 8y ago
  That would neither really restrict airflow to the point where AC would be feasible, or prevent passengers from falling onto the tracks. Not sure what that would solve as a problem.
  - danieltillett 8y ago
    The large heavy strip curtains would restrict airflow, especially ones that are attached top and bottom. Yes they wouldn't stop people falling on the tracks, but a least they could be installed without costing billions.
Terr_ 8y ago
One option is to increase the distance between the train and the doors, so that there's enough of a "vestibule" for people to move around in when a mismatch occurs.
- mislankanova 8y ago
  I have seen platform doors arranged like what you described at an inter-city railway station. Unfortunately this is unlikely to work for hot and gross subway systems like the NYC subway, where platforms tend to be narrow and overcrowded.
Gravityloss 8y ago
Every time there's some cool thing shown somewhere in the world, you have people telling excuses why it can't be done where they live. Close to the NIH synfrome but not quite the same. There are some valid concerns, but I often feel the first reaction tends to be quite negative. Maybe it doesn't make a comment people feel would be worth writing to say "great, let's do that here too!"
c22 8y ago
Why not just have the doors a few feet away from the train?
- pintxo 8y ago
  People will end up in this no-mans-land with the train gone, and the doors closed?
- bobthepanda 8y ago
  In many places there isn't "a few feet" available from the platform.

amarshall 8y ago

The biggest problem is indeed cost. Installing platform doors on the platforms throughout the system has been estimated in the past at US$1 million per station (US$400–500 million) in 2012 [1] and US$1 billion in 2013 [2]. But those estimates seem incredibly low now given more recent per-station costs elsewhere (CA$10 million, €2.6 million) [3], and the fact that NYC subway construction is higher than most other major cities [4]. As for depth, most NYC subway stations are directly below the street. I’d be surprised if NYC builds platform doors before Communications-Based Train Control (CBTC). [1] https://www.wnyc.org/story/283792-nyc-mta-tepidly-explores-platform-barriers-after-subway-track-deaths-again/ https://www.wnyc.org/story/283792-nyc-mta-tepidly-explores-p... [2] https://www.nytimes.com/2013/01/29/nyregion/mta-says-subway-platform-gates-could-cost-1-billion.html https://www.nytimes.com/2013/01/29/nyregion/mta-says-subway-... [3] https://www.thetransportpolitic.com/2017/09/26/the-case-of-the-missing-platform-doors/ https://www.thetransportpolitic.com/2017/09/26/the-case-of-t... [4] https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-construction-costs.html https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...

scarejunba 8y ago
But that's probably because NYC pays some multiple of what anyone else would pay. https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-construction-costs.html https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...
- amarshall 8y ago
  Yes, and I mentioned that (and the same article) in my post :). But even if you ignore NYC being more expensive, the more recent estimated per-station prices in Paris and Montreal would come to around US$1.5–3 billion for all of NYC’s stations. If NYC remains around 6x the price in Paris, then that’s around US$9 billion for platform doors in NYC.