NYCT - Bottlenecks Discussion Thread

[RR503]

2 hours ago, NBTA said:

Where can I find graphs like this?

I made them with a code a friend sent me that uses R and files downloaded from the MTA's public data feeds. If you're interested in me sending you a copy, dm me. 

[Deucey]

On 11/5/2019 at 11:35 PM, RR503 said:

Dunno about that. The data I'm looking at for the  show a solid 2-2.5 min median PM peak runtime premium from Canal to Dekalb. When the *median* loss is that much, yeah, you're inconveniencing people. Given the interchangability of the 6th and Bway corridors, the fact that the only people who'd really lose full dual access are folks at Brighton express stops and 36 St, and that there may be people who want the corridor they don't have, I'd def say this is worth considering. 

Also the reasons @Around the Horn and @T to Dyre Avenue give. Capacity and schedule reliability are important!

Would it be simpler to build a go-between tunnel in Manhattan that puts Bk bound 6 Av and Broadway trains on the south side of the Bridge and Uptown trains on the north side of the bridge?

[BreeddekalbL]

With current conditions,would it even be feasable to deinterline broadway today?

Edited December 12, 2019 by BreeddekalbL

[darkstar8983]

9 minutes ago, BreeddekalbL said:

With current conditions,would it even be feasable to deinterline broadway today?

it depends on which manner to de-interline. Also, I believe there are too few subway cars to provide the service necessary under any de-interlined pattern. 

[BreeddekalbL]

2 minutes ago, darkstar8983 said:

it depends on which manner to de-interline. Also, I believe there are too few subway cars to provide the service necessary under any de-interlined pattern. 

as in

as is

Astoria to whitehall but every 3rd train is sent to bay ridge on weekdays but weekends stays in manhattan

Coney island to 96/2

as is

[LGA Link N Train]

2 hours ago, BreeddekalbL said:

as in

as is

Astoria to whitehall but every 3rd train is sent to bay ridge on weekdays but weekends stays in manhattan

Coney island to 96/2

as is

I’m assuming that the number of Trains per hour would go as follows:

 on a combined 24 TPH

 15-16 TPH given Astoria’s terminal layout

 7-8 TPH

[BreeddekalbL]

1 minute ago, LaGuardia Link N Tra said:

I’m assuming that the number of Trains per hour would go as follows:

 on a combined 24 TPH

 15-16 TPH given Astoria’s terminal layout

 7-8 TPH

for that that would include the variable of every third train going south of whitehall?

[LGA Link N Train]

49 minutes ago, BreeddekalbL said:

for that that would include the variable of every third train going south of whitehall?

Now assuming that  service is boosted to 15-16 TPH and the  is running at 7-8 TPH, assuming that we’re working with a 15/7 ratio when it comes to Broadway Local service, 5  trains will go to Bay Ridge total. 

I’ll do a timetable later 

[BreeddekalbL]

17 minutes ago, LaGuardia Link N Tra said:

Now assuming that  service is boosted to 15-16 TPH and the  is running at 7-8 TPH, assuming that we’re working with a 15/7 ratio when it comes to Broadway Local service, 5  trains will go to Bay Ridge total. 

I’ll do a timetable later 

what would an 11/11 variable do?

[darkstar8983]

18 minutes ago, LaGuardia Link N Tra said:

Now assuming that  service is boosted to 15-16 TPH and the  is running at 7-8 TPH, assuming that we’re working with a 15/7 ratio when it comes to Broadway Local service, 5  trains will go to Bay Ridge total. 

I’ll do a timetable later 

or keep the current idea you have, but just flip the and  south terminals. The  is already the longest local service, and would be ideal to chop some of the route off. The Astoria-Bay Ridge route (yes it doesn't have a direct yard and needs to dead-head to Coney Island Yard, but its runtime would only be about 1 hour 15 minutes (better than 90 minutes)

 

- Astoria-Ditmars Blvd to Bay Ridge-95 St all times

Rush Hour Frequency: 16TPH

Midday Frequency: 8TPH

Evening Frequency: 10TPH

Weekend Frequency: 6TPH (likely will be scaled down to 5TPH due to constant construction)

- All times except late Nights: Forest Hills-71 Av to Whitehall St

Rush Hour Frequency: 8TPH

Midday & Evening Frequency: 6TPH

Weekend Frequency: 5TPH

Edited December 12, 2019 by darkstar8983

[LGA Link N Train]

1 hour ago, BreeddekalbL said:

what would an 11/11 variable do?

That’s 2 less  trains that would be sent to Bay Ridge. 11+5= 15 so that’s overall, 5 less  trains being ran in place of ... 5 extra  trains.

Edited December 12, 2019 by LaGuardia Link N Tra

[T to Dyre Avenue]

6 hours ago, BreeddekalbL said:

as in

as is

Astoria to whitehall but every 3rd train is sent to bay ridge on weekdays but weekends stays in manhattan

Coney island to 96/2

as is

You’d need to run far more  trains to sufficiently serve Astoria without the . As a matter of fact, they should be doing this. But like @darkstar8983 posted, it would be better to run the  to Bay Ridge and truncate the  at Whitehall. I don’t think Whitehall would be able to turn two out of every three ’s, especially if it were to be bumped up to 15-16 tph. 

2 hours ago, darkstar8983 said:

or keep the current idea you have, but just flip the and  south terminals. The  is already the longest local service, and would be ideal to chop some of the route off. The Astoria-Bay Ridge route (yes it doesn't have a direct yard and needs to dead-head to Coney Island Yard, but its runtime would only be about 1 hour 15 minutes (better than 90 minutes)

 

- Astoria-Ditmars Blvd to Bay Ridge-95 St all times

Rush Hour Frequency: 16TPH

Midday Frequency: 8TPH

Evening Frequency: 10TPH

Weekend Frequency: 6TPH (likely will be scaled down to 5TPH due to constant construction)

- All times except late Nights: Forest Hills-71 Av to Whitehall St

Rush Hour Frequency: 8TPH

Midday & Evening Frequency: 6TPH

Weekend Frequency: 5TPH

Right, and if the  runs at 8 tph during the rush, Whitehall should easily be able to handle it. As for where to store the , maybe a portion of 36th St Yard can be set aside to store  trains overnight and on weekends, while servicing the cars at CI (similar to the  with Fresh Pond Yard and ENY), while they figure out what to do with the entire 36-38 complex long term. 

Edited December 13, 2019 by T to Dyre Avenue

[RR503]

On 12/11/2019 at 10:47 PM, Deucey said:

Would it be simpler to build a go-between tunnel in Manhattan that puts Bk bound 6 Av and Broadway trains on the south side of the Bridge and Uptown trains on the north side of the bridge?

Not really -- track geometry isn't the biggest constraint in the area/the geometry constraints that do exist are mostly south of Dekalb Ave station (the curvy bits heading into the various Barclays platforms).

[RR503]

Thought y'all would enjoy this map:

https://drive.google.com/open?id=1Y-5FFDDOcrCR1YRK8wI3HEFK7bsc06_K

Edited December 17, 2019 by RR503

[JeremiahC99]

32 minutes ago, RR503 said:

Thought y'all would enjoy this map:

https://drive.google.com/open?id=1Y-5FFDDOcrCR1YRK8wI3HEFK7bsc06_K

That’s a great map. 
 

One thing I notice is that at Coney Island , all 4 lines are turning 10 trains per hour or less. Even if Brooklyn DeKalb Interlocking were deinterlined and the Broadway and other lines were also deinterlined to remove that bottleneck in 34th Street, what else is stopping them from turning 12+ trains per hour at Coney Island? I know it says on the map that many factors constrain train throughout, such as track infrastructure, but I wonder what makes Coney a bottleneck? Is it the switch layout for all lines? 
 

I’m asking since I see Coney Island as an integral part of my revised subway expansion (and deinterline) plan.

[RR503]

37 minutes ago, JeremiahC99 said:

That’s a great map. 
 

One thing I notice is that at Coney Island , all 4 lines are turning 10 trains per hour or less. Even if Brooklyn DeKalb Interlocking were deinterlined and the Broadway and other lines were also deinterlined to remove that bottleneck in 34th Street, what else is stopping them from turning 12+ trains per hour at Coney Island? I know it says on the map that many factors constrain train throughout, such as track infrastructure, but I wonder what makes Coney a bottleneck? Is it the switch layout for all lines? 
 

I’m asking since I see Coney Island as an integral part of my revised subway expansion (and deinterline) plan.

Coney Island definitely ranks high among the worst laid out terminals in the system. Switches are far from the terminal platforms (on the  and  especially), and the terminal is -- get this -- the last place automatic key by signals are in mainline revenue service use in the entire system.

Automatic key-by, for those of you who don't do signals, is a function whereby you can, in the simplest of terms, pass red signals by stopping *just* in front of the stop arm. Keying by was the villain in many a subway fender bender back in the day so was disabled on all mainline signals, but the layout of Coney Island's terminal platforms is such that signal engineers decided the most expedient way to ensure low speeds entering certain platforms was to install them. 

I'm frankly not sure of what the exact capacity of any given CI platform is, but I don't imagine it's high. 15tph or less, probably. Luckily there are _plenty_ of short turn locations available across BMT south. 

Edited December 17, 2019 by RR503

[Union Tpke]

4 minutes ago, RR503 said:

 

Automatic key-by, for those of you who don't do signals, is a function whereby you can, in the simplest of terms, pass red signals by stopping *just* in front of the stop arm. Keying by was the villain in many subway fender bender back in the day so was disabled on all mainline signals, but the layout of Coney Island's terminal platforms is such that signal engineers decided the most expedient way to ensure low speeds entering certain platforms was to install them. 

 

The obvious solution is to through run the  with the .  

[RR503]

Just now, Union Tpke said:

The obvious solution is to through run the  with the .  

That likely creates more operations problems than it solves -- the  and  added together would make for one _long_ route! I would just move some switches around, resignal the terminal and perhaps upgrade/reconfigure some short turn locations to support higher capacity and less disruptive terminal operations. 

[Union Tpke]

1 minute ago, RR503 said:

That likely creates more operations problems than it solves -- the  and  added together would make for one _long_ route! I would just move some switches around, resignal the terminal and perhaps upgrade/reconfigure some short turn locations to support higher capacity and less disruptive terminal operations. 

Which short turns would you fix? Kings Highway is being worked on now.

[Union Tpke]

3 minutes ago, RR503 said:

That likely creates more operations problems than it solves -- the  and  added together would make for one _long_ route! I would just move some switches around, resignal the terminal and perhaps upgrade/reconfigure some short turn locations to support higher capacity and less disruptive terminal operations. 

You could have your  from 96th Street to 96th Street (with deinterlining) and your  from Jamaica Center to Jamaica Center. 

[RR503]

Just now, Union Tpke said:

Which short turns would you fix? Kings Highway is being worked on now.

Depends on where we see demand growing/on just how bad each of the Stillwell terminals are/how bad we project relay ops to be in the future. First priority for me would be reconfiguring Brighton Beach so you can run a relay terminal off the local tracks w/o interfering w/  service, but you could also convert Bay Parkway/West End into a low-interference relay-friendly terminal by adding a switch between the s/b local and the middle just north of the station, or KHN to the same by adding a platform over one of the trackways, or....

Point being, there are options. 

3 minutes ago, Union Tpke said:

You could have your  from 96th Street to 96th Street (with deinterlining) and your  from Jamaica Center to Jamaica Center. 

LOL indeed you could. Though  from 96 to 96 would rely on Dekalb not being deinterlined. 

[JeremiahC99]

25 minutes ago, RR503 said:

Coney Island definitely ranks high among the worst laid out terminals in the system. Switches are far from the terminal platforms (on the  and  especially), and the terminal is -- get this -- the last place automatic key by signals are in mainline revenue service use in the entire system.

Automatic key-by, for those of you who don't do signals, is a function whereby you can, in the simplest of terms, pass red signals by stopping *just* in front of the stop arm. Keying by was the villain in many subway fender bender back in the day so was disabled on all mainline signals, but the layout of Coney Island's terminal platforms is such that signal engineers decided the most expedient way to ensure low speeds entering certain platforms was to install them. 

Well I wonder if its possible to move the switches on all the tracks closer to the station to get some more train throughput at that station. I'm looking at the track maps, and and it seems that all the switches are too far. On the  and , there should be double crossovers 20 feet north of the platforms for example so that they don't spend a lot of time wrong railing to/from the platforms. The ones at the bridge over the Coney Island Creek should be used for emergencies.

What else can be done to fix this?

25 minutes ago, RR503 said:

I'm frankly not sure of what the exact capacity of any given CI platform is, but I don't imagine it's high. 15tph or less, probably. Luckily there are _plenty_ of short turn locations available across BMT south. 

One of those locations should be Brighton Beach. It should handle more trains as well.

Also, would train throughput at the station increase if the express trains just pulled into the southbound platform, discharged all of its passengers on the platform, pulled into one of the two middle express tracks past the station, and reentered service on the northbound track (i.e, recreating Broad Street)? To me, the current situation where trains just pull into any track just recreates Flatbush on the , but on the . I would also like to see an increase of trains turned around there (rather than 10) for better service.

Edited December 17, 2019 by JeremiahC99

[RR503]

9 minutes ago, JeremiahC99 said:

Well I wonder if its possible to move the switches on all the tracks closer to the station to get some more train throughput at that station. I'm looking at the track maps, and and it seems that all the switches are too far. On the  and , there should be double crossovers 20 feet north of the platforms for example so that they don't spend a lot of time wrong railing to/from the platforms. The ones at the bridge should be used for emergencies.

What else can be done.

This would be my first line of attack -- moving crossovers. I'd also look carefully at the placement of tracks and crossovers at the other end of the platforms. The reason that we have AKs there is that, while there are no bumpers there except for the , a train that overruns the platform in any direction will end up in the middle of an interlocking, which is, ya know, not great. So to ensure trains will always stop before fouling switches, they have to enter quite slowly. If you can move the parts of the interlocking where the tail end of the terminal tracks merge with others, that'd be another good savings albeit one likely more difficult to achieve. 

11 minutes ago, JeremiahC99 said:

One of those locations should be Brighton Beach. It should handle more trains as well.

Also, would train throughput at the station increase if the express trains just pulled into the southbound platform, discharged all of its passengers on the platform, pulled into one of the two middle express tracks past the station, and reentered service on the northbound track (i.e, recreating Broad Street)? To me, the current situation where trains just pull into any track just recreates Flatbush on the , but on the . I would also like to see an increase of trains turned around there (rather than 10).

With current NYCT relay ops, you're probably better off as is with the qualification that Brighton Beach's crossover is a bit far from the platform. Relay terminals in this system are, thanks to long terminal dwells and the common coexistence of relays and DGTs, generally capable of only about 20tph, while our best turnbacks can do 30+. The issue with Flatbush, btw, isn't the platform config -- it's the fact that the tracks stub end, so trains have to enter slowly (as enforced by GTs) to ensure safety. 

Long term, even if we fix relay ops, I'd keep the  as is for two reasons: turnbacks are cheaper to operate than relays, and I would like to (given a reconfig) turn service at Brighton Beach off of the local tracks using the express tracks beyond the station). 

[JeremiahC99]

5 minutes ago, RR503 said:

Long term, even if we fix relay ops, I'd keep the  as is for two reasons: turnbacks are cheaper to operate than relays, and I would like to (given a reconfig) turn service at Brighton Beach off of the local tracks using the express tracks beyond the station). 

That is a great idea. I wonder how we reconfigure Brighton Beach to turn local trains at that station using the express tracks? Are you suggesting grade-separated junctions.

[RR503]

Just now, JeremiahC99 said:

That is a great idea. I wonder how we reconfigure Brighton Beach to turn local trains at that station using the express tracks? Are you suggesting grade-separated junctions.

Nah, just this. 

You'd be able to cross a train over to the express tracks, turn it, and then run back out without ever merging w/  service. Optimally we'd have some better, grade separated config, but this is what we've got to work with for now.