Second Avenue Subway Discussion

[KK 6 Ave Local]

On 5/2/2018 at 5:21 PM, bobtehpanda said:

For 6B in a normal country you could get to Fordham Plaza.

For 99 years of waiting (first proposal was in 1919) you could get to Boston.

[CenSin]

I remember making one before, but here’s a newly improved design for a 3-track terminal:

The design covers 3 scenarios:

1. The 3-track station is used as a terminal for the long term. The switches are arranged such that a train can leave the station while another can pull in to and from any combination of tracks. The stub tracks up north are for storing trains.
2. The 3-track station is used as a through station and short-turn terminal. The middle track is used for turning trains.
3. The 3-track station is only a through station. The middle track is not used for regular service.

Of course, there is a fourth scenario in which trains from the northern end use the station to turn southbound trains back north for a shuttle-like service.

For the situation where the station is used as a primary terminal, the switches are deliberately placed so that they sort out inbound traffic well before inbound trains get near the platforms, allowing for closely-spaced trains to consecutively charge into the station. The switches near the platform get outbound traffic out of the way immediately so that the track can be clear for incoming trains. This is key to turning trains at a fast clip. This should have the ability to turn 40 trains per hour easily.

[T to Dyre Avenue]

11 hours ago, CenSin said:

I remember making one before, but here’s a newly improved design for a 3-track terminal:

...

The design covers 3 scenarios:

1. The 3-track station is used as a terminal for the long term. The switches are arranged such that a train can leave the station while another can pull in to and from any combination of tracks. The stub tracks up north are for storing trains.
2. The 3-track station is used as a through station and short-turn terminal. The middle track is used for turning trains.
3. The 3-track station is only a through station. The middle track is not used for regular service.

Of course, there is a fourth scenario in which trains from the northern end use the station to turn southbound trains back north for a shuttle-like service.

For the situation where the station is used as a primary terminal, the switches are deliberately placed so that they sort out inbound traffic well before inbound trains get near the platforms, allowing for closely-spaced trains to consecutively charge into the station. The switches near the platform get outbound traffic out of the way immediately so that the track can be clear for incoming trains. This is key to turning trains at a fast clip. This should have the ability to turn 40 trains per hour easily.

Very well thought-out design. I like it a lot. It would be a hell of a lot better than the current setup at Main Street-Flushing on the , where it seems like only one train is allowed to enter or leave the station at a time, and slowly.

Edited May 12, 2018 by T to Dyre Avenue

[Porter]

 

Let's make this real simple and cheap for the (at least comparatively speaking).

Phase 2 (second and final phase): three new stations and two expanded stations:

<MANHATTAN> (Second Avenue Line)

106th Street
116th Street
125th Street

<BRONX> (Third Avenue Line)

138th Street
149th Street

Yes, I understand that ending the project here would defeat the purpose of a true 'Second Avenue' subway line, which I do agree is needed, but given the 's financial and operational ineptitude and misfortunes, I think this is the best that can be achieved in the near future. A serious Second Avenue Subway with new and lines will have to come sometime later down the line, perhaps a few decades at this rate. Therefore, I think we should focus on what can be reasonably achieved, which I've outlined above, a plan that reasonably accommodates future extensions along 125th Street in Manhattan and Third Avenue in the Bronx, not to mention further south along Second Avenue in Manhattan.

Since the can barely keep the existing infrastructure together, it should really just aim for a more reasonable goal like this one.

Edited May 12, 2018 by Porter

[CenSin]

7 hours ago, T to Dyre Avenue said:

Very well thought-out design. I like it a lot. It would be a hell of a lot better than the current setup at Main Street-Flushing on the , where it seems like only one train is allowed to enter or leave the station at a time, and slowly.

That’s what this is for (assuming the Flushing Line gets extended ever):

Just now, CenSin said:

Overkill solution for a high throughput line that’s triple-tracked at one end and double-tracked at the other:

The design could hypothetically support 60 trains per hour one way for a brief period until the yard at either end runs out of trains to send out or the local track in the other direction starts limiting throughput due to saturation.

[bobtehpanda]

Don't we already have high-capacity two-to-three terminals in the form of Parkchester and Whitehall? I feel like this is overengineering.

[MysteriousBtrain]

1 minute ago, bobtehpanda said:

Don't we already have high-capacity two-to-three terminals in the form of Parkchester and Whitehall? I feel like this is overengineering.

The track setups create terrible bottlenecks and create many delays. Bedford Park Blvd is kind of set similar and works well during the rush.

[CenSin]

8 hours ago, bobtehpanda said:

I feel like this is overengineering.

It depends on the specific project’s needs. And it says so right on the sticker:

15 hours ago, CenSin said:

Overkill solution for a high throughput line

I really don’t know what place could have a need for 60 trains per hour in the rush direction along one line when 2 separate parallel lines could do the same (and maintain the same throughput around the clock). Perhaps a nation-state is building a high-density planned city over some undeveloped government-owned land with cheap labor, and this somehow fits right into their built-from-scratch subway.

But such a design is certainly nice to have when the common constraints that limit us to sub-par solutions are lifted.

 

8 hours ago, bobtehpanda said:

Don't we already have high-capacity two-to-three terminals in the form of Parkchester and Whitehall?

Neither of those could really be considered high-capacity/high-throughput if you consider the number of tracks and/or the slow procedures needed to turn a train. They generally work by supplementing the line with additional trains which, combined, improves throughput. But taken alone, they themselves are not capable of high-capacity nor high-throughput operation. It also depends on your definition of high-capacity and high-throughput. Is 1 pocket track enough? Do we care about holding up trains in the tunnel because the terminal track is occupied? Is it alright to turn just 10 trains per hour?

Capacity:

• No capacity: a 2-track station with no crossover switches has 0 tracks for turning trains (14 Street–Union Square ); they are purely through stations.
• 1 track: a 3-track station with only switches to the middle track (Whitehall Street–South Ferry ) has 1 track for turning trains.
• 2 tracks: a 2-track station with 2 crossover switches (facing and trailing) has 2 tracks for turning trains (34 Street–Hudson Yards ).
• 3 tracks: a 3-track station with crossover switches connecting all 3 tracks has 3 tracks for turning trains (Flushing–Main Street ).

Whitehall Street–South Ferry can hardly be called high capacity as it’s the lowest ranked in terms of terminal capacity.

Throughput:

1. Terminating trains on the same track as those used by through service adds delays. (Fix: no terminating trains on through service tracks)
2. Fumigation adds delays. (Fix: altering the policy to not fumigate or by having operators at both ends of the train)
3. Waiting for available terminal track adds delays. (Fix: more terminal tracks)
4. Flat junctions add delays. A scissor crossover (or “diamond switch”) is essentially a flat junction in 50% of the cases. (Fix: flying junctions)
5. Low-speed switches add delays. (Fix: high-speed switches)

 

Most New York City subway terminals are limited in throughput primarily by the dearth of tracks for terminating trains (low capacity) and/or all of the above points concerning throughput. Of the 46 stations that have dedicated tracks for holding and turning back trains that are also in regular use, 35% have 1 track, 63% have 2 tracks, and 2% (only Flushing–Main Street) have 3 tracks. Stations that use through service tracks for terminating trains are  excluded from the figures—such as Church Avenue.

With terminal throughput, only #2 is easily fixable by altering policy which may not necessarily come with an increase in spending. #3, #4, and #5 being physical limitations practically means that the job has to be done right from scratch.

[bobtehpanda]

5 hours ago, CenSin said:

It depends on the specific project’s needs. And it says so right on the sticker:

I really don’t know what place could have a need for 60 trains per hour in the rush direction along one line when 2 separate parallel lines could do the same (and maintain the same throughput around the clock). Perhaps a nation-state is building a high-density planned city over some undeveloped government-owned land with cheap labor, and this somehow fits right into their built-from-scratch subway.

But such a design is certainly nice to have when the common constraints that limit us to sub-par solutions are lifted.

 

Neither of those could really be considered high-capacity/high-throughput if you consider the number of tracks and/or the slow procedures needed to turn a train. They generally work by supplementing the line with additional trains which, combined, improves throughput. But taken alone, they themselves are not capable of high-capacity nor high-throughput operation. It also depends on your definition of high-capacity and high-throughput. Is 1 pocket track enough? Do we care about holding up trains in the tunnel because the terminal track is occupied? Is it alright to turn just 10 trains per hour?

Capacity:

• No capacity: a 2-track station with no crossover switches has 0 tracks for turning trains (14 Street–Union Square ); they are purely through stations.
• 1 track: a 3-track station with only switches to the middle track (Whitehall Street–South Ferry ) has 1 track for turning trains.
• 2 tracks: a 2-track station with 2 crossover switches (facing and trailing) has 2 tracks for turning trains (34 Street–Hudson Yards ).
• 3 tracks: a 3-track station with crossover switches connecting all 3 tracks has 3 tracks for turning trains (Flushing–Main Street ).

Whitehall Street–South Ferry can hardly be called high capacity as it’s the lowest ranked in terms of terminal capacity.

Throughput:

1. Terminating trains on the same track as those used by through service adds delays. (Fix: no terminating trains on through service tracks)
2. Fumigation adds delays. (Fix: altering the policy to not fumigate or by having operators at both ends of the train)
3. Waiting for available terminal track adds delays. (Fix: more terminal tracks)
4. Flat junctions add delays. A scissor crossover (or “diamond switch”) is essentially a flat junction in 50% of the cases. (Fix: flying junctions)
5. Low-speed switches add delays. (Fix: high-speed switches)

 

Most New York City subway terminals are limited in throughput primarily by the dearth of tracks for terminating trains (low capacity) and/or all of the above points concerning throughput. Of the 46 stations that have dedicated tracks for holding and turning back trains that are also in regular use, 35% have 1 track, 63% have 2 tracks, and 2% (only Flushing–Main Street) have 3 tracks. Stations that use through service tracks for terminating trains are  excluded from the figures—such as Church Avenue.

With terminal throughput, only #2 is easily fixable by altering policy which may not necessarily come with an increase in spending. #3, #4, and #5 being physical limitations practically means that the job has to be done right from scratch.

60 trains per hour is not being achieved anywhere on Earth today on a two-track line. The highest capacity I have ever heard is 40TPH. Computer controlled, moving block trains still have to maintain braking distance between trains.

Stations that have two or three tracks on either side have two modes of operation - either all trains are passing through, half of the trains are passing through, or no trains are passing through. In the first case, terminal capacity is not relevant; in the second, you can deal with sub-standard amounts of capacity (Whitehall just needs an approach track and it would be fine for the purpose), and in the third a three track terminal is actually a detriment to capacity due to all the conflict movements, the inability to have consistent downtown-oriented tracks, etc. It's telling that 34-HY is going to be the highest-capacity terminal for a two-track line in town once Flushing CBTC is used to its fullest potential, which is because it both has tail tracks (allowing trains to enter at speed) and has crossovers immediately after the station, so there's no need to relay. And that's just for two tracks.

[CenSin]

14 minutes ago, bobtehpanda said:

60 trains per hour is not being achieved anywhere on Earth today on a two-track line. The highest capacity I have ever heard is 40TPH. Computer controlled, moving block trains still have to maintain braking distance between trains.

You’re reading the diagrams wrong. For the 2-track diagram, I specifically mentioned 40 trains per hour: post #5052 at the very bottom. The most recent diagram is for a 3-track line, which is capable of 30 train per hour per track for two tracks operating traffic in the same direction. But since you mentioned CBTC, perhaps my conservative estimate should actually be 80 train per hour. Of course, there is a limit to how long this can be sustained on an asymmetrical setup where only 1 track returns trains from the other end, but that’s also been acknowledged.

19 minutes ago, bobtehpanda said:

a three track terminal is actually a detriment to capacity due to all the conflict movements, the inability to have consistent downtown-oriented tracks, etc.

That’s exactly what the track diagram addresses—removing conflicting movements.

21 minutes ago, bobtehpanda said:

It's telling that 34-HY is going to be the highest-capacity terminal for a two-track line in town once Flushing CBTC is used to its fullest potential, which is because it both has tail tracks (allowing trains to enter at speed) and has crossovers immediately after the station, so there's no need to relay. And that's just for two tracks.

It’s good that the MTA is actually building these terminals with capacity in mind. But theoretically, traffic will still not have as high of a throughput as if the scissor crossovers had been a flying junction instead.

There is also a scenario that is often ignored: trains waiting in the tunnel for terminal tracks to be available. Having had both Flushing–Main Street and Coney Island–Stillwell Avenue as destinations regularly, I’ve often found myself waiting on trains outside the terminal station because there is no terminal track available. (The situation is a lot better in Flushing, owing to the quick clip at which trains leave the terminal and the number of terminal tracks available.) The MTA has definitely thought about three-track terminals since they’ve shown up in plans for 125 Street () and 34 Street–Hudson Yards (), but these are the first major elements of terminal designs to be shelved when costs go up.

[R68OnBroadway]

How feasible is it to add express tracks to 2nd (or side tracks) but on the outer ends on the platforms (the express-local setup seen on the red line on the Chicago L near Howard)? Doing this allows for greater capacity and increased service along the line, and could permit the  to run to 125th and then run crosstown while the  and  could operate to the Bronx and have an express-local setup there along Third or Webster. 

You could also eliminate the express service addition but still have outer tracks at 72nd so you could send the  there and still have 2 terminating tracks for it.

[CenSin]

1 hour ago, R68OnBroadway said:

How feasible is it to add express tracks to 2nd (or side tracks) but on the outer ends on the platforms (the express-local setup seen on the red line on the Chicago L near Howard)? Doing this allows for greater capacity and increased service along the line, and could permit the  to run to 125th and then run crosstown while the  and  could operate to the Bronx and have an express-local setup there along Third or Webster. 

You could also eliminate the express service addition but still have outer tracks at 72nd so you could send the  there and still have 2 terminating tracks for it.

How would outer tracks be added? The entire cavern is under an arching roof (which prevents its collapse). An independent cavern or bore could be made—under the existing cavern or adjacent to it at the same level, but underneath buildings (because the current cavern is entirely underneath all of the roadway already).

[R68OnBroadway]

13 minutes ago, CenSin said:

How would outer tracks be added? The entire cavern is under an arching roof (which prevents its collapse). An independent cavern or bore could be made—under the existing cavern or adjacent to it at the same level, but underneath buildings (because the current cavern is entirely underneath all of the roadway already).

Outer tracks would be adjacent to the current tracks, but you could build a lower level if that is better.

[CenSin]

2 minutes ago, R68OnBroadway said:

Outer tracks would be adjacent to the current tracks, but you could build a lower level if that is better.

There’s going to be some uncomfortable level of separation between the two, mainly because the walls can’t just be knocked down to open up an expanded mezzanine for additional tracks along the side.

[R68OnBroadway]

3 minutes ago, CenSin said:

There’s going to be some uncomfortable level of separation between the two, mainly because the walls can’t just be knocked down to open up an expanded mezzanine for additional tracks along the side.

Would 2-tracked lower level terminal for the  work better?

[bobtehpanda]

On 5/17/2018 at 5:25 PM, R68OnBroadway said:

Would 2-tracked lower level terminal for the  work better?

It'd probably be easier to build. We converted 59th St on the Lex while it was an active subway stop.

Realistically speaking by the time the bypass happens, 72 should be built as four-track with a new lower level. 

Edited May 19, 2018 by bobtehpanda

[Wallyhorse]

2 hours ago, bobtehpanda said:

It'd probably be easier to build. We converted 59th St on the Lex while it was an active subway stop.

Realistically speaking by the time the bypass happens, 72 should be built as four-track with a new lower level. 

And very similar to what I would do, except with an additional stop on 79th Street at York-1st Avenues with that three tracks to allow for short-turning there.

[CenSin]

49 minutes ago, Wallyhorse said:

And very similar to what I would do, except with an additional stop on 79th Street at York-1st Avenues with that three tracks to allow for short-turning there.

That’s an odd place for a short-turn; it’s after the 63 Street tracks merge with the lower level tracks. For smoother operations, it would make more sense to build the lower level of 72 Street exactly as the original 72 Street station was intended the first time around and have the 63 Street tracks connect to the middle tracks at 72 Street lower level to stay out of the way of through traffic.

[bobtehpanda]

4 hours ago, CenSin said:

That’s an odd place for a short-turn; it’s after the 63 Street tracks merge with the lower level tracks. For smoother operations, it would make more sense to build the lower level of 72 Street exactly as the original 72 Street station was intended the first time around and have the 63 Street tracks connect to the middle tracks at 72 Street lower level to stay out of the way of through traffic.

How does that result in 'smoother operations'? With a lower level 72nd you can turn around basically the full Broadway express while still having through SAS service. That would not have been the case with the three-track 72 St.

[CenSin]

5 hours ago, bobtehpanda said:

How does that result in 'smoother operations'? With a lower level 72nd you can turn around basically the full Broadway express while still having through SAS service. That would not have been the case with the three-track 72 St.

You have connections from the southern segment of 2 Avenue to the 79 Street connector. That would imply the possibility of regular service from 2 Avenue to Queens via that connector.

[Lawrence St]

6 hours ago, bobtehpanda said:

How does that result in 'smoother operations'? With a lower level 72nd you can turn around basically the full Broadway express while still having through SAS service. That would not have been the case with the three-track 72 St.

72nd St was suppose to be a three track station from the start, but why they changed it is beyond me.

It seems to me MTA likes to build extensions fast and take out some of the good original blueprints and modify them to their liking, THEN when they finish the extension is when they realize they made a mistake and want to either add a new station (i.e 11th Av ) or modify the station later (adding a lower level at 72nd St).

 

[Eric B]

They did it to save money. (How much money is beyond me).

[CenSin]

5 minutes ago, Eric B said:

They did it to save money. (How much money is beyond me).

Hey, we got a spacious mezzanine and some art out of it.

[bobtehpanda]

12 hours ago, CenSin said:

You have connections from the southern segment of 2 Avenue to the 79 Street connector. That would imply the possibility of regular service from 2 Avenue to Queens via that connector.

Sure. In the event of a service change you don't necessarily need to keep that service pattern going, however.

There has always been a huge issue with how 63rd St's junction was designed. In its original form, you had three pairs of tracks leading into two through pairs and one terminal.

• IND 63rd -> Queens
• BMT 63rd -> one terminal track
• SAS -> SAS

No matter how much juggling you do one trunk line is going to have a cap on capacity to/from the south. Building out an entirely new second level, and then eventually another connection to Queens, mitigates that by having three track pairs going north or east.

It's why the current setup of Queens service makes no sense; there are three pairs of tracks leading to three pairs of tracks, but to maximize one seat rides we just leave a sixth of cross-Manhattan capacity on the table.

[CenSin]

This was what I meant:

Broadway trains can turn using the lower level middle track. There is no interference with any other services. Broadway trains can also go to Queens, and 2 Avenue trains can turn using the lower level middle track. Again there is no interference with any other services.

Edited May 21, 2018 by CenSin