NYTIP – extending the nyc subway, part 1: sas phase 2 and laguardia airport

UPDATE (03.16.2020): Post substantially revised to include four new options for SAS Phase 2!

Welcome to my ongoing NYTIP series! In the next series of posts, I will expound on point 2 in my three-point plan to fix the NYC Subway – extend. I’ll begin with Phase 2 of the Second Avenue Subway (SAS) and rail service to LaGuardia Airport. The current plans fall short, so allow me to discuss a better way.

NOTE: Click any image to enlarge.

I. SAS Phase 2

The current plan extends the Q train past its 96th Street terminus to Lexington Avenue – 125th Street station, connecting to the 4/5/6 trains and Metro-North. Curiously, the plan calls for tail tracks extending past Malcolm X Blvd – 125th Street station (2/3 trains), but no connecting station. The plan retains a provision for future service to The Bronx. Unfortunately, this phase isn’t scheduled to open until at least 2027.

That’s where NYTIP comes in.

Previously, I discussed the Broadway and Queens Boulevard de-interlining, where the N and Q trains serve 63rd Street/SAS and the R train serves Astoria. With both the N and Q trains serving SAS, NYTIP proposes two SAS extensions – one across Harlem and one to The Bronx. Under NYTIP, both extensions comprise a single phase, which I’ll call SAS Phase 2X. In SAS Phase 2X, the N serves The Bronx and the Q serves the cross-Harlem branch.

Let’s explore the many ways we can realize SAS Phase 2X.

Option 1a: Regional Plan Association (RPA) recommendation – Cross-Harlem and Concourse extensions

[Fig. 1] Overview of Option 1a with emphasis on the Cross-Harlem (125th Street) branch of SAS Phase 2X, served by the Q train. Created using Brand New Subway.

[Fig. 2] Overview of Option 1a with emphasis on the Concourse extension, served by the N train. SAS via Concourse requires some changes to A/C service.

The RPA proposed two extensions of SAS Phase 2 in their Save our Subways publication. RPA’s SAS Phase 2B is the Cross-Harlem extension, and SAS Phase 2C is the Concourse extension. Under NYTIP, the Q serves the Cross-Harlem extension and the N serves the Concourse extension. The N serves new stations at 3rd Avenue – 138th Street (6 train), 149th Street – Grand Concourse (2, 4, and 5 trains), and 161st Street – Grand Concourse (4 and B trains), then meets the Concourse line at a point south of 167th Street station.

To provide some relief on the overcrowded 4 line, the N would run via Concourse express in the peak direction, while the B runs local. Since the Concourse line is only three tracks, NYTIP recommends running a maximum of two services on this branch. Thus, the C – originally rerouted via Concourse as part of the CPW de-interlining – would run north to Inwood instead. Owing to the limitations of the existing 207th Street terminal on the A line (an island platform with no tail tracks), Option 1a includes a new terminal for the C train just east of the 207th Street Yard leads.

[Fig. 3] C extension to Inwood via 207th Street Yard leads. The original track map belongs to vanshnookenraggen.

NYTIP contemplates a terminal station with side platforms for several reasons. First, the leads are immediately below street level. Second, 207th Street runs downgrade, leaving no room for a mezzanine – a requirement for an island platform. Finally, the station site is close to the Harlem River. Importantly, this short extension can also serve as a provision for a Bronx crosstown subway via Fordham Road.

The downside to this option – aside from the capital investment required – is less service at the 155th Street – 8th Avenue station on the Concourse line. An alternate option exists to address both downsides.

Option 1b: Cross-Harlem – Concourse connection

[Fig. 4] Overview of Option 1b.

Option 1b is a modified version of Option 1a that requires less new construction, albeit at the expense of a direct route from Concourse to the SAS. A key advantage of this option over Option 1a is doubled service on the Cross-Harlem line at Lexington Avenue, Malcolm X Blvd, and St. Nicholas Avenue, and increased service at 145th Street station. As with Option 1a, the A and C trains serve Inwood and the B and N trains serve The Bronx, with the B via local and the N via peak express. New track connections preserve the CPW de-interlining, as follows:

[Fig. 5] Track connections between the SAS and the Concourse line via CPW stubs.

[Fig. 6] Overview of the CPW-SAS connection using CPW stubs.

Even in the absence of a Concourse-SAS service, the track connection shown in Fig. 6 would allow operational flexibility and SAS access to both the 207th Street and Concourse yards. Given that both the 207th Street and Concourse yards are potential storage/maintenance facilities for SAS, this track connection is crucial. In my view, this makes Option 1b more desirable than Option 1a.

But there are other options for extending the SAS to The Bronx.

Option 2: 3rd Avenue Subway

[Figs. 7, 8] Overview of Option 2.

Option 2 is the oft-discussed 3rd Avenue subway. Under NYTIP, the N train would serve this line. It is the long-overdue replacement for the 3rd Avenue El, which MTA demolished in 1973. It closes a long gap between the Concourse line and the southern portion of the White Plains Rd line. The relatively high ridership of the existing bus services – the Bx15 and Bx41 (respectively, the 19th and 21st-busiest routes in NYC) – make the 3rd Avenue subway a strong contender.

The Regional Plan Association once touted the 3rd Avenue SAS branch. In recent years, however, RPA changed their Bronx-SAS plan to SAS Phase 2C. (See Option 1a above.) RPA’s justification for this is their T-REX proposal, where enhanced local service on Metro-North’s Harlem line serves the 3rd Avenue corridor. As someone who lived on or near the Grand Concourse for most of my life, I can appreciate RPA trying to breathe new life into the oft-neglected Concourse line. On the other hand, the T-REX proposal is very intriguing. It has its flaws, but integrating regional rail is key to enhancing New York-area transportation. I will devote a future series of posts to regional rail under NYTIP.

Option 3: Morris Park subway

[Fig. 9] Overview of Option 3.

Option 3 begins the same way as Option 2, with a subway serving 3rd Avenue – 138th Street, The Hub – 149th Street, and 163rd Street stations along 3rd Avenue. Rather than continuing north, the line takes a northeasterly route via Boston Road, Tremont Avenue, and Eastchester Road to Morris Park. The purpose of this route is to relieve the overcrowded 2 and 5 trains, as well as enable a future transit hub at Morris Park – the site of one of four new Metro-North stations comprising Penn Station Access. Importantly, the FRA is also contemplating regional rail service at Morris Park (see section 3.7.6 of the linked document). A new rail yard could be build in the vicinity of the Hutchinson River Parkway and I-95 due northeast of the proposed terminal.

Option 4: Lafayette Avenue subway

[Fig. 10] Overview of Option 4.

Option 4 is my take on the proposed Lafayette Avenue subway – a plan dating back to 1939. As with Options 2 and 3, the line serves three stations on 3rd Avenue. Under Option 4, the line curves eastward south of 163rd Street, so the proposed 163rd Street station is east-west rather than north-south like in Options 2 and 3. The line runs via 163rd Street, Hunts Point Avenue, Lafayette Avenue, and Tremont Avenue to Throggs Neck. Unlike Options 2 and 3 which are wholly underground, NYTIP contemplates a mostly-elevated line for Option 4 due to two water crossings (Bronx River and Westchester Creek). The likeliest location for an el portal is Lafayette Avenue east of Hunts Point Avenue, given its width. A rail yard could be built in the vicinity of the Bruckner Interchange.

One thing you’ll note is that all options include the Cross-Harlem extension. This is a recommended extension under NYTIP, regardless of which Bronx extension is ultimately built. Importantly, choosing one option doesn’t necessarily preclude the others if the SAS is ever expanded beyond two tracks (future post).

Recommended path forward: Option 2. The 3rd Avenue subway connects to a major transit hub at Fordham Plaza and closes a large gap between subway lines in The Bronx.

Alternate recommendation: Option 1b. If funding precludes a full Bronx extension, then the Cross-Harlem – Concourse connection offers the biggest bang for the capital buck.

II. LaGuardia Airport

LaGuardia Airport is one the busiest airports in the United States, yet it has no direct rail link. It is widely criticized as a substandard airport in terms of flight delays, outdated facilities, hellish traffic, and the aforementioned lack of rapid transit options.

A few years back, Governor Cuomo unveiled an ambitious redevelopment plan for the airport, which includes an AirTrain. However, this AirTrain is rife with problems, from its proposed routing away from Manhattan, to its $2 billion price tag, to its Flushing Bay alignment. Mind you, the project was initially slated to cost $450 million and was supposed to run over the Grand Central Parkway (GCP).

It’s clear we need something different.

The previous plan for a LaGuardia rail link called for extending the Astoria El to the airport – a plan that dates back to 1943! Community opposition killed this plan each time the City proposed it. However, this extension is the only sensible way to bring rail to LaGuardia Airport, making it a key element of NYTIP.

[Fig. 11] Astoria line extension to LaGuardia Airport, served by the R train.

Given that the environmental review process for the proposed LaGuardia AirTrain just started, there is still a chance to kill the AirTrain boondoggle and pursue a superior rail link. NYTIP contemplates an Astoria El extension that combines elements of EIS Alternatives 7 and 8 – elevated from Ditmars Blvd to 45th Street and 19th Avenue, then underground from there to LaGuardia Airport. The extension includes new stations at Steinway Street, Hazen Street, 94th Street – LaGuardia Airport West, and East Elmhurst – LaGuardia Airport East; it also includes a new storage yard within the Con Edison property north of the Ditmars Blvd station. While EIS Alternatives 7 and 8 call for the Astoria El extension to follow the GCP to LaGuardia, the proposed route under NYTIP follows the GCP only up to 92nd Street. East of 92nd Street, the route runs via Ditmars Blvd. This alignment retains connectivity to LaGuardia Airport while providing stations closer to the surrounding neighborhood.

Since the R does not merge with any other line under NYTIP, an optional service enhancement is peak-direction Astoria express service. The previous iteration of this service was unpopular and hence short-lived; however, the LaGuardia extension and the service increases made possible by NYTIP could make it viable again. Even if the R made all stops at all times, it’d still be faster than the “backwards” LaGuardia AirTrain!

In my next post, I’ll discuss the Concourse line extension. Until next time!

13 thoughts on “NYTIP – extending the nyc subway, part 1: sas phase 2 and laguardia airport

  1. This is a good plan for a northern extension of 2 Ave. Curious to see if a southern extension of 2 Ave can work. By my estimation, it would be hard to fit in a 2 Ave south of 63rd street and still maintain good deinterlining.*

    Of course if there are two separate northern terminals for 2 Ave trains (125/Broadway and Fordham Plaza), this would mean that all of the trains of the Broadway BMT express must go to 2 Ave (as you show) and there is no room to divert any Broadway BMT expresses to Queens Blvd (as I suggested in a comment to address Queens Blvd crowding that would not be successfully diverted to the G train). This would mean then diverting some of the Broadway BMT locals onto Queens Blvd instead of Astoria (which will not be possible if the Astoria service is more popular with LGA service) or biting the bullet and making capital improvements to the 8th Ave local southern terminal so that all Queens Blvd service can reach Manhattan without utilizing the Broadway BMT at all. [You had suggested a Worth Street subway, others have suggested perhaps a Spring Street subway to divert some 8 Ave trains onto the Williamsburg Bridge – that will also have the effect of allowing more 6 Ave trains to service the Rutgers tunnel and Culver line. Regardless, all of these tunnels are significant captial expenses that must be done, even prior to any extension of 2 Ave to two separate northern terminals.]

    In that case, yes, all Broadway expresses to 2 Ave. All Broadway locals to Astoria and LGA. All Queens Blvd trains to utilize the 53 and 63 tunnels to service 8 Ave and 6 Ave respectively.

    * One idea is that Q goes from Broadway to upper 2 Ave. N goes from Broadway to 63rd street to Queens. T goes all the way on 2 Ave. V goes from 2 Ave to 63 street to Queens. Not deinterlined, but service to many locations is maintained. It does mean, though, that there will be less service to the 8 Ave local and 6 Ave local as both of those will be serviced through the 53 st tunnel. So if this were a long range plan, there is no need to address the capacity at WTC along 8 Ave.

    Liked by 1 person

    1. I’m curious about the Spring Street alignment. I know there was discussion on Twitter about splitting the J and hooking SAS to Williamsburg (though I don’t think it’s the same plan as the one you mentioned).

      For SAS under NYTIP, what I’m thinking is having the full-length SAS run express north of 63rd Street on separate tunnels. This ensures no conflicts with the Broadway services that run local. This also does not preclude an SAS-Queens connection; if the mainline SAS operates express along its length, the SAS-Queens services could run via SAS local, maximizing capacity. This is all subject to change of course, but that’s where I’m leaning.


      1. Can’t find it now, but there was a map where someone essentially drew a connecor along Spring Street from 6 Ave to Chinatown. The idea is that it will provide a second terminal for 8 Ave trains and allow for full service along the 8 Ave line and it will connect to the willy bridge line a little west of the Bowery station. Not an official plan by any means.

        Ideally, a deinterlining plan can work to an extent because there are 6 northern portals, 6 central lines, and 6 southern portals. They are not currently all connected in the best way, but a few key connections could work:

        Upper 2 Ave – 2 Ave – Broadway Express – Man Br S – Brighton
        Astoria – 60 St – Broadway local – Montague Tunnel – Bay Ridge

        W Hts/Bronx – CPW local – 6 Ave Express – Man Br N – 4 Ave Express to Sea Bch/W. End lines
        QB local – 63 St – 6 Ave local – Rutgers tunnel – Culver Line*

        Inwood/Bronx – CPW express – 8 Ave Express – Cranberry Tunnel – Fulton line
        QB express – 53 St – 8 Ave local – NEW – Will. Bridge – Myrtle/Jamaica*

        For something like this to work, there are several things that are needed. Yard access for the Astoria line. Something to address the way that G interlines with the F in Brooklyn so that G will no longer interfere with the F. A subway connection along Spring Street from the 8 Ave line to the Williamsburg Bridge. Maybe some work on the ramps near the W4 St station that allow 6 Ave trains to head towards WTC and 8 Ave trains to head east on Houston St – something i refer to as the w4 switcheroo. Perhaps making the 36 St station on QB line an express stop. New in-system transfers 59/Lex to 63/Lex, Prince to Bwy-Lafayette, Q plaza to QB Plaza, 53/7 Ave to 57/7 Ave

        * QB is the hardest to deinterline even with dedicated access to the 53 and 63 stations. With the placement of the 11 St cut as access to the local lines, it is best to have the QB locals use 63. It is also true that 53 is a busier part of midtown, so it needs express trains. The problem is that the stations east of the Willl Bridge cannot take the longer trains, so we want to get the QB locals on the Will Bridge and the QB expresses onto the Culver line. This means using the W4 switcheroo. In that scenario:

        QB express – 53 St – 8 Ave local – W4 switcheroo – Rutgers tunnel- Culver line
        QB local – 63 St – 6 Ave local – W4 switcheroo – WTC (if no construction, so limited nubmer of trains)
        OR Will Bridge (limited size of trains due to old stations)


      2. Thinking about this more, that Spring Street connector and W4 switcheroo might work better than Worth Street. The reason why I didn’t consider W4 initially is because the Chrystie connector diverges east of B’way-Lafayette, creating a new conflict there. A Spring Street connector avoids this problem by preserving connectivity to Williamsburg; running 8th Ave service via Rutgers/Culver avoids the WTC issue.

        So you could have a modified version of my QB Option 3 long-term, except the E/K take B-Laf and Culver, while the F takes WTC (shortening its route and eliminating the super-long-local issue) and the M takes Spring-Willy B, preserving its route in Brooklyn.

        The opportunity cost of the Spring Street connector is fixing limitations on J/M service imposed by the capacity-killing at-grade merge at Myrtle. One thought I had long-term was rerouting the M via a rebuilt Myrtle El/subway and having it join the Rutgers line at Jay St, eliminating this merge and allowing massive service increases (including a full peak-direction express) on the J.

        The G interlining is indeed a challenge – one I don’t have a solution for as yet.


  2. I have thought of other deinterlining ideas that may account for trains better. The idea is to find what works with minimal capital investment. To work best, there still has to be yard access for the Astoria line.

    Upper 2 Ave – 2 Ave – Broadway Express – Man Br S – 4 Ave express to West End/Sea Beach

    Broadway express is completely deinterlined.

    Astoria – 60 St – Broadway local – Montague Tunnel – Bay Ridge [a new Astoria yard required]

    Broadway local is completely deinterlined.

    Inwood/Bronx – CPW express– 6 Ave Express – Man Br N – Brighton lines

    6 Av express is completely deinterlined.

    QB local – 63 St – 6 Ave local – Rutgers tunnel – Culver Line or Myrtle Ave line

    6 Ave local, north of Broadway-Lafayette is deinterlined. South of there, it runs the existing service of the F and M trains. F will interact with G. M will interact with J. So while the 6 Ave locals interact with other trains, there will be no interaction with any other trunk lines (Broadway ex/local, 8th Ave ex/local, 6th Ave ex).

    QB express – 53 St – 8 Ave Express – Cranberry Tunnel – Fulton line
    W Hts/Bronx- CPW local – 8 Ave local – WTC

    The above can work, but it’s a little out of the box. Switches do allow for 53rd st trains to be 8th Ave express, (and all 8th Ave trains can serve 50 St). Both Queens express and Fulton are very heavy trains so their frequencies do match up. (Plus, each service serves three possible terminals: Fulton trains service Fulton locals to Euclid, Fulton expresses to Lefferts, and Fulton expresses to Far Rockaway; QB expresses service QB express to FH-local to 179, QB express to 179, and QB express to Jamaica Center.)
    It is a very long routing, but the routing avoids the WTC terminal for Queens trains.

    WTC is not a problem for CPW locals as those have a far lower demand than Queens.

    And the trains sort of fit like a process of elimination. If Queens express go to 8th Ave express, then CPW locals must service the 8th Ave local. Then, the CPW expresses must service 6th Ave express, and the Queens locals must service the 6th Ave locals.

    So long as Astoria-Bay Ridge is served by a yard, it seems like we have a significant deinterlining system wide. – without capital expenditure other than the yard. Plus, all Queens Blvd trains service Manhattan (no G train) without impacting Broadway.

    I hope to address the problem with deinterlining if we don’t get a yard for Astoria tomorrow.


    1. For the CPW de-interlining, you certainly could choose 6th Avenue express and 8th Avenue local, which would replace the C/D swap I proposed with an A/B swap. Although QB and Fulton have heavier demand than the CPW local, the WTC terminus is still a limiting factor for the total number of trains you can run. Consequently, if you split A/C service between Washington Heights and The Bronx, each would be infrequent even during rush hours. I don’t know the maximum TPH limit for WTC, but I don’t think it’s higher than 20 TPH – and it may be lower. Thus, this pattern limits service at the local stops in Washington Heights and the Concourse to an absolute maximum of 10 TPH peak. You’d still need some form of relief (e.g. stub tracks using the Worth Street subway provision, or the Worth Street subway itself, or that Spring Street connector) to allow further service increases.

      Streamlining the Lefferts/Rockaway split the way you propose here is an interesting idea. While your proposed routes are long, express service offsets the length. Consider that once upon a time, the MTA operated the rush-hour C local from Bedford Park to Rockaway Park! Also, the E used to be an 8th Avenue express train via Fulton Street decades ago.


  3. My apologies, the last comment didn’t forward well with formatting, hopefully this will work:

    Without a yard to service Astoria-Bay Ridge line, the full deinterlining that I suggested in my last comment would not work. Every service other than the 6 Ave expresses, Broadway expresses, and Broadway locals can run as mentioned above. If there is no yard, there are 5 main options in routing the services through DeKalb to South Brooklyn.

    For all options, B and D are CPW-6 Ave expresses (and thus have access to yards in Northern Manhattan and/or Bronx); N and Q are Broadway expresses; and R and W are Broadway locals. W trains will all go from Astoria to Whitehall, although to the extent that R trains also originate from Astoria in most of the options, a W may not be necessary, as it can be labeled as a short-run R.


    B Brighton exp
    D West End
    N Astoria-Bwy exp-Sea Beach
    Q 2 Ave – Bwy exp – Brighton local
    R Qns loc 60 – Bwy loc – Bay Ridge
    W Ast – Bwy loc – Whitehall

    Everyone knows this is horrible as N trains are forced to switch between locals and express. Plus there is a mess at DeKalb.

    Trains serving DeKalb: B,Q,R
    Trains serving the 4 Ave/Pacific platform at Atlantic (cross-platform): D, N, R


    B Brighton exp
    D W End [or Sea Bch]
    N Qns loc 63-Bwy exp-Bay Ridge
    Q 2 Ave – Bwy exp – Sea Bch [or W End]
    R Ast- Bwy loc – Brighton loc
    W Ast – Bwy loc – Whitehall

    As with the other options presented here, the solution to the yard problem is to send the Astoria-Bwy locals to
    either Whitehall or the Brighton local. The idea is that every branch to Coney Island (WE, SB, Brighton) will still have at least one train leading to the Manhattan Bridge. Send the Astoria-Bwy local-Tunnel trains through as Brighton locals to access CI yard. Thus, Bay Ridge trains along the 4 Ave local will have access to either the 6 Ave expresses (yards in Northern Manhattan/Bronx) or the Broadway expresses (with access to Queens Blvd). Of course, to the extent that Queens Blvd connects to Broadway there is interlining. All of the options also involve some level of interlining at De Kalb.

    This and all other options presented prevent Bwy locals and Bwy expresses from interfering with each other in Manhattan. That is probably the most critical thing that the MTA should do in the short term.

    Trains serving DeKalb: B,R, N
    Trains serving the 4 Ave/Pacific platform at Atlantic (cross-platform): D, Q, N

    One negative with this and the other options is that the W End and Sea Bch will have to make two cross-platform transfers (Atlantic and DeKalb) to reach the R train for access to Lower Manhattan.

    OPTION 3

    Option 3 is like Option 2, except we isolate the 6 Ave trains from interference at De Kalb. Bwy locals and Bwy expresses still interline at De Kalb.

    B W End
    D Sea Bch
    N Qns loc 63-Bwy exp-Bay Ridge
    Q 2 Ave – Bwy exp – Brighton exp
    R Ast- Bwy loc – Brighton loc
    W Ast – Bwy loc – Whitehall

    Trains serving DeKalb: Q,R, N
    Trains serving the 4 Ave/Pacific platform at Atlantic (cross-platform): B, D, N

    Sea Beach and West End passengers have direct access to 6 Ave express and a cross-platform access at Atlantic to the Broadway express.

    OPTION 4

    Option 4 is like Option 2, except we send the Bay Ridge trains to 6 Ave instead of the Bwy express. This will allow us to separate Queens Blvd from the Broadway trains.

    B Bay Ridge
    D W End [or Sea Bch]
    N 2 Ave -Bwy exp-Sea Bch [or W End]
    Q 2 Ave – Bwy exp – Brighton exp
    R Ast- Bwy loc – Brighton loc
    W Ast – Bwy loc – Whitehall

    Trains serving DeKalb: Q,R, B
    Trains serving the 4 Ave/Pacific platform at Atlantic (cross-platform): D, N, B

    OPTION 5 (My preference)

    Option 3 is like Option 4, except we isolate the Bwy exp trains from interference at De Kalb. Bwy locals and 6 Ave expresses still interline at De Kalb.

    B Bay Ridge
    D Brighton exp
    N 2 Ave – Bwy exp – Sea Bch
    Q 2 Ave – Bwy exp – W End
    R Ast- Bwy loc – Brighton loc
    W Ast – Bwy loc – Whitehall

    Trains serving DeKalb: D,R, B
    Trains serving the 4 Ave/Pacific platform at Atlantic (cross-platform): N, Q, B

    Sea Beach and West End passengers have direct access to Bwy express and a cross-platform access at Atlantic to the 6 Ave express.

    This option is my preference as it isolates Bwy express and keep Qns Blvd away from Bwy. 6 Ave expresses still interline with Bwy locals to provide two trains to Brighton and one train to Bay Ridge so that all trains have access to some yard. (The interlining is no worse than what we currently experience at DeKalb.)

    These are options that can be implemented without capital expenditure. Of course, the best option, if there was a yard in Asotria, is to send the Bay Ridge trains to Astoria and thus fully deinterline DeKalb.


    1. I saw that you revised your comment to fix the formatting, so I went ahead and deleted the duplicate comment with the bad formatting.

      Obviously, the status quo that ties things up at the Gold Street interlocking has to change. Interestingly, for Options 2 through 5, you could run the R via Brighton express if you wanted to without affecting yard access. I’d imagine that, if most Brighton riders prefer Broadway service as they did when the Manny B reopened in 2004, this might not be the best option. But it is an option nonetheless.

      Options 2 and 4 retain the Gold Street knot and are hence undesirable. Options 3 and 5 work better. To make them even simpler, you don’t need to swap the B and the D under either of those. For Option 3, you could still have D via West End (the B would serve Sea Beach); for Option 5, the B could remain as the Brighton express and the D could serve Bay Ridge. I agree that, absent the Astoria yard, some form of Option 5 would be best as this does not preclude a full Queens Blvd de-interlining. Also, whatever serves Bay Ridge under Option 5 does not conflict with Broadway thanks to the switch layout s/o DeKalb, making it preferable to the jumbled mess we have now.


  4. Regarding the 8 Ave locals as CPW locals to WTC: I see your point and it is true that WTC is a limiting factor. To actually get some hard numbers on the maximum ability to turn around trains would be extremely helpful.

    I’m under the working assumption that the mid-section of a track, with proper signaling and perhaps a limitation on the use of timers, can run 30 TPH (every 2 minutes) easily. Most of the terminals can probably handle within the 15-20 TPH range, which means that the usual practice of sending two trains down the straightaway tracks, split into two separate terminals on either end works well [e.g. sending B and D down the 4 Av express in Brooklyn, but having only B to Sea Beach and D to West End.]

    Three way splits are not ideal, but some already exist in the system. The trains in the Cranberry tunnel are already regularly split into C to Euclid, A to Lefferts, A to Far Rockaway, and even one or two A’s to Rockaway Park. [I really wish the MTA would use a different letter to distinguish the Rockway service from Lefferts service, especially that many tourists use the train to JFK. They shouldn’t get lost at Lefferts.] The Queens Blvd express does that too, although to be truthful there are only a handful of expresses to 179 and those are only utilized becuase of the 12 TPH limitation at Jamaica Center. That being said, it is possible that some of the Concourse locals can be service to an extent by 6 Ave expresses. FWIW, at least in Midtown there is no substantial difference between the 8th Ave express and the 8th Ave local, since if the E were to run express the only stations that are skipped by the express are 23rd and Spring. Again, my idea was to optimize the system ASAP [just a few months time to rework schedules and update maps] without the need for capital construction where there is no money.

    Regarding the South Brooklyn trains:

    Any option pushing the Broadway locals onto tracks other than the 4 Ave local would result in a protest by the local population. I felt that pushing the Broadway local to one of the Brighton trains would mean that there would still be the other Brighton train with access to the Manhattan Bridge. In my mind it seemed fairer than to push the only train on the Sea Beach or the West End to the Montague tunnel, but a similar type analysis could work there. If R trains (Astoria-Bwylocal-Montague tunnel) went to Sea Beach, there would be less intermixing, since all of the West End trains already split off earlier, but the switches at 4Av/59 probably can’t handle the heavy mixing. If R trains went to West End, they could take advantage of the flyover junction that is there at 4 Av/36, but the problem is [other than the protests from West End riders] that there are no switches south of the junction so that some of the 4 Ave expresses can serve the local 4 Ave stations south of 36th. If the B and D ran on the 4 Ave express, the D should continue to Sea Beach and the B should serve Bay Ridge, but to do that it should also serve the stations between 36 and 59, but it cannot without extra track switches. Regardless, nobody along Brighton, West End, or Sea Beach would be happy to trade their access to the Bridge for an R train through lower Manhattan, even if it ran express between Coney Island and DeKalb.

    The only other option that may work would be an extension of a Nassau Street subway service to Bay Ridge. Prior to the cancellation of the V train, the M train during rush hours ran through the Montague tunnel to Bay Pkwy along the West End line. What if this service were recreated (and extended to more parts of the day, except late nights), but sending M trains to Bay Ridge and R trains to Bay Pkwy? [Or if the M’s stay on 6 Ave, extending J service to Bay Ridge.] To be conservative with the MTA’s resources, I imagine that this operation would only send 6 R’s and 6 M’s each hour through the tunnel. Most trains on either line will terminate somewhere in Lower Manhattan: City Hall and Whitehall for the R/W and Chambers and Broad for the J/M/Z. This would put 12 TPH (every 5 minutes) on the 4 Ave local north of 36 St and provide service to both sides of Lower Manhattan (granted Nassau and Church are not that far away from each other). The service to Bay Ridge would be somewhat paltry at 6 TPH, but it wouldn’t be crowded as most people along the corridor will switch to an express as soon as they could. Such a plan would separate 6 Ave and Bwy expresses at DeKalb, and only introduce a small amount of interlining at the Montague tunnel and at 36 St as R locals merge in with whatever express trains service the West End line.

    Again, a yard in Astoria would solve all of these issues, but this seems like a viable interim measure, particularly as it would allow a majority of the system to operate more efficiently.

    It is particularly sad that service to Bay Ridge was actually better when the Montague tunnel was closed due to Sandy repairs. This is because of the poor effects of interlining and the cascading delays that the R experinecing due to all the intermixing that it does, especially along Queens Blvd and the N switching from local to express at Herald Square. An M train to Bay Ridge would probably be as effective for Bay Ridge riders as the Brooklyn only shuttle, and it does provide a yard at the northern end of the line.


    1. Had a few new ideas that may also bring some of this together. Consider:

      $ Inwood/Conc exp – CPW exp – 8 Ave exp – Cranberry – Fulton lines – Ful loc to Euclid/ Ful exp to Leffert/ Ful exp to Far Rockaway [A, C, H trains]

      [I think this was your idea, so not much to talk about. Current A train with some variants. Unified line from 145th to Hoyt/Schermerhor running at full capacity. Reasonable intermingling at 145th]

      $ Jam Ctr/ FH to 179 loc/ FH to 179 exp – QB exp – 53 – 8 Ave loc – W4 swicheroo – (1/3 of trains terminating at 2 Av/Houston) – Rutgers tunnel – Culver line [E, blue V, blue K trains]

      [Here, I’m deinterlining the M train away from 6 Ave. J/M/Z to stay on Nassau Street-Williamsburg Bridge and not interfere with any other lines. Terminate some trains at 2 Ave to allow for room for the G train to come in. Some intermingling with G train, but no intermingling with other trunk lines. This means that Rutgers tunnel is only operating at 2/3 capacity, but the QB/8th Ave segment north of 2 Ave/Houston to Forest Hills running at full capacity. There is a possibility that some Culver trains will terminate at Kings Hwy and that there will be room for some Culver expresses, but that is not a priority for me. The point is that since all trains stay on IND tracks, they can be run full-lenth.]

      $ Forest Hills – QB loc – 63 St – 6 Ave loc – W4 switcheroo – WTC [F train]

      [Full-length trains to service this segment, but limited to WTC’s terminal capacity. The difference in capacity between FH and WTC to be addressed by some G trains that will interline along the QB loc. Some intermingling with G train, but no intermingling with other trunk lines. By my estimation, it means that 63 tunnel and 6 Ave local operate at 16 TPH.]

      $ W Hts/Conc loc – CPW loc – 6 Ave exp – M Br N – Brighton local / 4 Ave local to Bay Ridge [B, D train]

      [As discussed in earlier comment, De Kalb intermingling with Bwy locals in order to provide Bay Ridge with access to Concourse yard. Also reasonable intermingling at 145th. And some intermingling at 4 Av/36 st in Brooklyn, see below. I address your concern about providing enough CPW loc service by briging the CPW locals to Brooklyn and not terminating at WTC. ]

      $ Upper 2 Ave – Bwy exp – M Br S – 4 Ave express – Sea Beach / West End [N, Q train]

      [This line is fully de-interlined, except with some intermingling at 4 Av/36 St in Brooklyn, see below.]

      $ Astoria – 60 St – Bwy loc – Montague tunnel – Brighton exp / 4 Ave local to West End [R, W train]

      [De Kalb interingling with 6 Ave express trains. Some intermingling at 4 Av/36 St in Brooklyn.]

      A slight variation from my earlier comment addressing bringing 6 Ave trains to Bay Ridge. Instead of terminating all W trains at Whiethall, what if we brought them some through the Montague tunnel to provide additional service through the tunnel.

      We are recreating the old pattern of intermingling at 4 Ave/ 36 St where one 4 Ave local goes to Bay Ridge, one 4 Ave express goes to Sea Beach, and a local and express both service West End, with one train going to Bay Pkwy and one train going to CI. IMO, ideally we can also provide some express service on the West End line itself. Intermingling is never ideal, but sometimes necessary to provide more access to different locations, and just as done at 145th street, provide as many people with access to the Bridge and Manhattan expresses. There is still a lot less interlining than current service. The intermingling in this case does allow for West End express service that could speed up a lot of southern Brooklyn trips to Midtown.

      B trains servicing Brighton local. Every Brighton line station has direct access to the Bridge and an express train to Midtown via 6 Ave. Access to Upper Manhattan and CI yards. Transfer to R at any Brighton express station for access to Lower Manhattan. Transfer to 6 train at Bwy/Lafayette for access to any area btwn City Hall, Canal Street, and Union Square in lieu of access to Bwy express trains. [A free transfer to Prince St BMT would also help for access to eastern Manhattan below 23rd Street.]

      D trains servicing Bay Ridge. Every 4 Ave station has direct access to the Bridge and an express train to Midtown via 6 Ave. Access to Concourse Yards. Transfer to R at DeKalb or W at any station from 36 St north for access to Lower Manhattan. Transfer to N,Q trains at any 4 Ave express station for access to Bwy express.

      N trains servicing Sea Beach. Every Sea Beach staion has direct access to the Bridge and an express train to Midtown via Broadway. Access to CI yards. Cross-platform transfer at Barclay’s to D trains to 6 Ave. Cross-platform transfer at Barclay’s to W trains to Lower Manhattan or backtrack with a transfer at Canal Street to R/W, 6 , or J/Z.

      Q trains servicing West End express. Q trains should start at Bay Pkwy and provide express service on the West End line to further speed up trips in southern Brooklyn. Most West End stations will not have direct service to the Bridge, but for those at the sourthern end of the line, the access to the express will save them time. Passengers on the Q train boarding at a West End express station will have direct access to the Bridge and an express train to Midtown via Broadway. Access to CI yards. Cross-platform transfer at Barclay’s to D trains to 6 Ave. Cross-platform transfer at Barclay’s to W trains to Lower Manhattan or backtrack with a transfer at Canal Street to R/W, 6 , or J/Z.

      R trains servicing Brighton express. This is to be marketed as an express train to the Financial District, the only one servicing southern Brooklyn. Transfer from B to R train to get quickly to De Kalb, Downtown Brooklyn, and Lower Manhattan. Transfer to B/D at DeKalb for express service to Midtown via 6 Ave. Transfer to B/D trains at De Kalb and transfer again at Bwy/Lafayette for 6 train access to eastern Manhattan below 23rd street in lieu of Bwy express trains. Access to CI yards.

      W trains servicing West End locals. While most West End stations will no longer have direct access to the Bridge, their overall trip is made quicker with the ability to transfer to Q trains at the Bay Pkwy, 62nd, and 9 Ave. For passengers on the W trains, a transfer to the Q will provide direct direct access to the bridge and a Broadway express. Transfers at any 4 Ave station to the D train to access 6 Ave express service. W trains provide access to Downtown Brooklyn and Lower Manhattan. Access to CI yard.

      A lot of related discussion of this can be found on this thread:


  5. A quick separate word on train frequency. I assume 30 TPH as the ideal frequency to provide a train every 2 minute for a track with upgraded signaling and no interlining. But where it is not ideal, 24 TPH is probably a good working number. It assumes a train every 2.5 minutes plus, 24 is a very nice mathematical number as it is evenly divisible by 2, 3, and 4.

    A 24 TPH can be split into two train services.

    AB-AB, 50-50 split. Each line at 12 TPH or every 5 minutes.

    If an uneven split is better, the following option are also workable:

    AAB-AAB. A trains at 16TPH, B trains at 8 TPH. B trains every 7.5 minutes. A trains at 2.5 or 5 minute intervals.
    AAAB-AAAB. A trains at 18 TPH, B trains at 6 TPH. B trains every 10 minutes. A trains at 2.5 or 5 minute intervals.

    A 24 TPH can also be split into three train servcies

    ABC-ABC. Even split. Each line at 8 TPH or every 7.5 minutes.
    ABAC-ABAC. A trains at 12 TPH, every 5 minutes. B and C trains at 6 TPH or every 10 minutes.
    ABABC-ABABC. If we assume 25 TPH, this allows for A and B trains at 10 TPH each and C trains at 5 TPH. A C train approximately every 12 minutes, and A and B trains at approximately every 5 minutes.

    When dividing up a line into two or three branches, especially when deinterlined, there is no need to divide up the line evenly. I imagine that some of the services that split will be better to be split in an uneven fashion as above. For example, the three-way split of Fulton trains will probably be best as an ABABC pattern, rather than a ABC pattern. I imagine that Fulton locals to Euclid and Fulton expresses to Far Rockway have far more traffic than Fulton expresses to Lefferts. With only 3 unique stations, Lefferts trains can probably survive with half the service of Fulton expresses to FR. Local trains, though, should probably have about as many trains as the FR expresses.


    1. Hi mrsman: sorry for the long delay in responding.

      All of these ideas are interesting and worthy of consideration, but it’s important not to lose the plot here. The focus of NYTIP is to improve service in the most efficient manner possible – in general, simpler is better. There are certainly many possible ways to achieve this (and I still have a few more posts to update before continuing this series), but I’m not going to go over every possible iteration. In general, I pursue service patterns that minimize uneven splits. Fulton is an outlier for the reasons you mentioned – one branch (Rockaway) is much longer than the other (Lefferts). Here, I pursue the long game – have SAS take over the Fulton local, allowing A/C express (A to Rockaways, splitting between Far Rock and Rock Park – eliminating the shuttle – and C to Lefferts). Once you go uneven, you increase the chances of irregularity, which as you know ripples throughout the system.

      For Culver, even pursuing the W4 swap, you still have to deal with the G (especially if you send two services through Rutgers). Clearly, one of the two Rutgers routes would go express in that scenario to minimize interlining, but you could also either truncate the G to Hoyt (not popular), or send it somewhere else (but again, that’s the long game); for now, the G mixes in like in the QB partial deinterlining option.


Discuss respectfully. No trolling or threats allowed - violators subject to moderation or ban. Thanks!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s