Regional Rail for New York, Part II

In my previous post, I presented a proposal for half-hourly diesel-electric regional rail service in New York, to be implemented after completion of Gateway, East Side Access, Penn Station Access, and the Third Track. With all the additional capacity, providing this kind of robust regional express capacity would be possible. However, with time, the excess capacity would be exhausted. In addition, it would not fix one of the most pressing problems of the transportation network; currently, passengers are dumped into the local transit system at just two locations, overloading the subway system. If capacity on the rail network is reached after the completion of the initial phase of work, a main goal should thus be to more evenly distribute passengers, to relieve east-west and north-south subway lines. The goal is something akin to London’s Crossrail, or the Paris RER; a cross-city link with multiple stations providing an alternative express network within the city.

That being said, there are several caveats to building such a network. First off, the small width of Manhattan makes building multiple stations in a straight east-west direction very difficult; in the studies for East Side Access, a East Side station on the existing rail line serving Penn was ruled out as too difficult, expensive, and disruptive to build, and was considered operationally undesirable since all trains would have to stop at this new station. To the south, the main sites for potential stations, WTC and Fulton, are so close together that they are connected by a walking passageway, and any such east-west station would have to be carefully nestled into the thicket of lines around Lower Manhattan.

Connecting the future Penn South to GCT also wouldn’t do a lot by itself; passengers heading downtown would still overload north-south subway lines. In addition, if inbound capacity into either terminal was reached during the AM Peak (or outbound in the PM peak), it wouldn’t do much to provide additional peak capacity; it would certainly improve equipment utilization, but ultimately the tracks are the main restricting factor in running more trains, and in the case of Grand Central and Metro-North there are more than enough tracks and platforms to accommodate all the trains that could terminate there, even with inefficient practices. Linking Penn South with the LIRR’s cavern at GCT wouldn’t be good in the long-term either; through-running trains would have to fit in the tight clearances of the 63 St tunnel, so trains with multi-voltage transformers or double-deck cars would not be able to through-run, ruling out the possibility of through-running on track without third rail.

The electric fleets for the three operators would further limit possibilities for connections. As previously mentioned, they all run on different power supplies; making an all-in-one EMU would be very difficult considering the three AC catenary systems and two DC third rail systems. The current fleets were all bought at different times as well, and are sizeable; replacing every last train with a new, multi-voltage train would take a long time, holding up implementation of electric regional rail until most or all of the trains necessary were available.

The solution I am proposing would thus be more similar to Melbourne’s City Loop, Sydney’s City Circle, and New York’s Nassau Street Loop. The primary component of this proposal would be two new two-track tunnels; one for the LIRR from Grand Central to Atlantic Terminal via the East Side and Lower Manhattan, and another two-track tunnel for NJT running from Penn South to Hoboken via the East Side and Greenwich Village. The two tunnels would meet at a central transfer station at Union Square, providing direct regional rail access to fast-growing areas of Brooklyn via the L train.

Click for full image.

Click for full image.


The two tunnels central to the second phase of regional rail.

Greenwich Village Stop

A Greenwich Village stop could potentially be built along Houston or King St.

Intermediate stops at Greenwich Village and at Fulton St would provide additional access to Manhattan and also allow for additional transfers to the subway system at Houston St on the 1 and Spring St on the C and E trains, and also provides new east-west transit links, redistributing passengers more efficiently and reducing the load on the subway.

In New Jersey, the former Morris & Essex Lines (Gladstone Branch, Morristown Line, and Montclair-Boonton Line) would only serve the tunnels via Hoboken, while the Northeast Corridor, Raritan Valley Line, and North Jersey Coast line would only serve the tunnels via Penn. This segregation would allow for a doubling of capacity on those lines while reserving the original two North River Tunnels for use on the diesel regional network and Amtrak.

The proposed configuration of New Jersey Transit services through the new tunnel.

The proposed configuration of New Jersey Transit services through the new tunnel.

In the east, the Port Washington and Hempstead branches, along with a new branch to a park & ride at Belmont Park, would run via GCT and Atlantic to Far Rockaway, Long Beach, and West Hempstead. This would be the basis of a new metro-style express service between the inner neighborhoods of Nassau and the City, and super-frequent services would allow for a few additional stops to be added within the city itself without significantly impacting overall travel times, boosting intra-city as well as suburban transportation.

  • On the Port Washington branch, two additional stops at Elmhurst (Queens Blvd) and Corona (National St) would provide service to an area that currently relies on the overcrowded Q58 and 7 train.
  • The Main Line would see an additional station at Rego Park where the former Rockaway Beach Branch split off.
  • On the Atlantic Branch, an additional stop at Woodhaven Blvd would be built.
  • On the Far Rockaway and Long Beach branches, stops would be added at Linden (Linden Blvd) and Rochdale (Baisley Blvd).
  • On the West Hempstead branch, stops would be added at Farmers (Farmers Blvd) and Springfield (Springfield Blvd).

Upon electrification of the Oyster Bay Branch, services from Belmont Park could be shifted to the Oyster Bay Branch, with diesel services shifted over to the other branches of the regional rail system. Alternatively, Oyster Bay and Babylon trains could also run through the tunnel upon electrification, but this would split the tunnel amongst four services instead of three, limiting frequencies on the outer branches to six trains an hour, or a train every ten minutes; still a large increase in service.

Proposed LIRR "Metro-East" service, with new stations shown.

Proposed LIRR “Metro-East” service, with new stations shown.

With all this service redirected out of the main railroad tunnel to Penn, express services to Penn on the New Haven Line and Hudson Line could be redirected via Penn to provide additional capacity in the event of the Park Avenue tunnels reaching capacity. Long-term, the success of these plans would also depend on the streamlining and grade separation of junctions throughout the system, and the continued progress of the installation of new signalling, double-tracking, and electrification. However, if fully realized, the plans would result in the biggest change in regional-area transit since the construction of the subway.

Posted in Uncategorized
6 comments on “Regional Rail for New York, Part II
  1. Alon Levy says:

    Ugh, I only saw this now. This is what I get for not reading other blogs more regularly. (Also, I could’ve sworn I put your blog on my blogroll, but it isn’t there right now.)

    Anyway: I think there are a lot of problems with both versions of the proposal (the medium-term diesel option, and this long-term option), but I’ll put my responses in this post only, to consolidate.

    The diesel option suffers from assuming a lot of concrete and no electronics. ESA is a $10 billion project. Gateway as currently proposed is $16 billion. That’s a lot of money to spend on tunnels for 130-ton diesel locos. It’s also a long enough investment that, by then, the M7s will need mid-life refurbishing, and NJ Transit will have bought many new trains.

    Diesel locomotives suck. In electric mode they suck a bit less, but they still suck. They’re noisy and heavy and accelerate very slowly.

    You’re partly counteracting the slow acceleration problem by having the system use the farther-out, more express branches of the LIRR, but this is exactly backward. Through-running is of the greatest benefit to inner-suburban riders. If you live in Queens, you can plausibly work in Newark given through-running. If you live east of Ronkonkoma and take a train, your commute is hell no matter what, so you’re probably suffering it for a job that only exists in Manhattan.

    All of this seems to center around the need to avoid reelectrifying a few tens of kilometers of 12 kV 25 Hz track with 25 kV 60 Hz. This cost is a rounding error, possibly even lower than the cost of building a loop to let trains from the Erie lines go to Penn Station. It’s worth reelectrifying the entire NEC at 25 kV purely for performance, since it would allow the newer trains to be lighter. If I remember correctly, a recent German transformer would be 4.5 t at 16.7 Hz, 1.25 t at 60 Hz, and 3 t at 25 Hz, so the difference is 1.75 t per car, which is 3-4% of train mass.

    The long-term all-electric option reminds me of the oldest versions of my regional rail proposal. (The bit about having trains loop back, Nassau Street-style, was proposed by Wdobner in comments, and I still think it’s a bad idea.) The Jersey waterfront station is Hoboken, and not Pavonia or Exchange Place, even though they’re located in secondary CBDs and Hoboken isn’t, and Pavonia can be reached from Secaucus on existing infrastructure.

    Similarly, the station between the Jersey waterfront and Lower Manhattan is in the West Village station, which in isolation would be a good idea, but is not the best place to site a station between the Jersey waterfront and Lower Manhattan if you can vary your routes. Taking a wider arc to hit West 4th, or going via Pavonia and then stopping at Chinatown, both hit more important nodes. I discussed some of these tradeoffs here.

    Finally, you’re underrating the importance of a Penn-GCT connection. Grand Central is near the middle of the Midtown CBD; Penn Station is at its edge. Under your plan, commuters from New Jersey can connect to the 4/5/6 at Union Square, but it’s circuitous, and doesn’t add too much value over taking the E from Penn Station. The connection from New Jersey to Lower Manhattan requires a transfer at Union Square as well, and is again circuitous.

    • Henry says:

      Gateway or an equivalent to it has to be built to provide additional capacity (and provide backup in the event of a North River Tunnel shutdown). East Side Access already has the station and tunnels carved out, and is a sunk cost at this point regardless of whether or not the project made any sense in hindsight. I don’t believe that M7s would be refurbished to the extent that they get overhauled with multi-voltage systems, especially considering the clearance issues with the 63 St tunnel, and NJ Transit’s prospects for either further electrification or radically different trains are not something I consider particularly good considering the current funding situation of that particular agency. (The MTA has similar issues, but the agency has been tottering back and forth on money issues for quite a while now.) Even if NJT were to reelectrify everything to 25 kV 60 Hz, the only line east of Penn Station that can handle that is the Northeast Corridor; unless we already have third rail shoes that can switch between over-running and under-running without going to the shop, NJT would need to buy separate fleets for through-running on MNR and LIRR.

      I’m proposing diesel-electric expresses instead of through-running electric regional rail primarily because of the layout of New York itself. Manhattan has been and continues to be an overwhelmingly dominant job center; the region has no suburban office areas on the scale of La Defense or Canary Wharf. Queens residents who want to go to New Jersey job centers can transfer to a regional express at Sunnyside; people in Brooklyn can change at Union Square; and everyone who wants to go upstate can change at GCT, Penn, or Sunnyside. If transfers on this system are cross platform or a simple escalator ride up or down, then it’s definitely preferable to schlepping to the E if you’re riding all the way to Midtown East. Connecting Penn to GCT and through-running would be contingent on reelectrifying the NE Corridor south of Penn, and it’s not as if Grand Central suffers from a lack of platforms or tracks to clear trains from.

      If by Chinatown you mean Canal Street, I’m not a particularly big fan of the idea of a station there. Canal St as it is today is a traffic sewer choked with congestion, and isn’t the center of Chinatown, much less a potential business or shopping district to place a regional station at. The subway stations there are also barely connected by a warren of twisting and turning routes across platforms in a mess that is arguably worse than the former Fulton Street, so building a station connecting to the N/Q/R/J/Z/6 may not be a good idea. The other potential station location on Canal, to connect to the 1 and the A/C/E, is too close to the Holland Tunnel ramps to be a viable location to promote development. West 4th has issues as well; the B/D/F/M don’t actually go downtown, the streets are very irregular, and connecting to the 1 would be difficult considering the proximity of both stations to the PATH tubes. There’s also the issue of potentially impacting Washington Square Park itself; in a neighborhood full of NIMBYs, it is a treasured public space, and fights to preserve the park haven’t been lost, even against major institutions like NYU. Jersey is the part of the metro area I am familiar with, but the basic concept of a loop is still something I’d prefer for distribution purposes.

      • Alon Levy says:

        ESA is a sunk cost, but Gateway is not. It’s a tunnel of many billions of dollars. Reelectrifying 100 km is at most in the low hundreds of millions – the catenary’s already there, and although 25 kV requires a few inches more clearance, only the tunnels themselves are in a constrained location, and even they have enough space.

        There is no real need for any NJT train to be equipped with underrunning shoes. The trains would go to Long Island, where they’d use overrunning shoes, or the NEC, where they’d keep their catenary. The only reasonable through-running that requires two kinds of third rail is LIRR-to-Empire, and that line needs to be catenarified anyway because of intercity rail needs. (Yes, it has nontrivial cost, but it’s a second-order term by the standards of how much new tunnels cost.) Even with a Penn-GCT connection, it’s better to put catenary in for a few km than to get dual-voltage trains.

        As for the line-matching and station pattern, first, La Defense wasn’t what it is today when the RER A opened. Second, I think of West 4th or Chinatown as a neighborhood stop and not a CBD stop. There’s more street activity at Broadway and Canal than at Nation. (Canal also connects to Williamsburg easily, but Union Square does even better, and Canal’s connection to Downtown Brooklyn via the N/Q is devalued given a direct commuter rail link.)

        Third and most importantly, Nassau Street-style systems don’t do distribution well. The original intent of the Nassau Street Loop was to get people to Downtown Brooklyn and Lower Manhattan and waste less time turning at terminals; connecting to other trains was not part of the plan. Two-seat rides are fine, but the loop system still adds an unnecessary transfer and several minutes for most riders. Consider how on social media people are yelling at me for proposing to do away with LIRR Main Line express trains at rush hour, adding 5 stops and 6 minutes to the trip from Hicksville to Jamaica, even without the extra transfer. (I think those people are insane, but 5 minutes is not the same as 5 minutes plus a transfer.)

      • Henry says:

        I don’t think Nassau St is a particularly good comparison due to the relative tightness of the loop; a better example would be the Melbourne or Sydney rail systems, which make wider loops through the center for additional capacity. Granted, I’m not entirely sure if they function well, but it’s certainly not guaranteed to be a failure. The orange M makes an almost loop between Queens Blvd and Ridgewood, and it’s doing well. It does force transfers for certain trips, but those are trips that already have transfers, and the loops are meant to function in conjunction with the diesel-electric regional rail, which does through-run into New Jersey.

        I personally prefer Greenwich Village since that’s about as close as you can get to NYU without disrupting the university, Washington Square Park, or the Houston St subway, and currently Union Sq-Greenwich Village is a somewhat difficult crosstown trip to make. (Union Square – Canal St is a trip you can already make using the N/Q/R.) While it would improve access to New Jersey, the current conditions at Canal St (unsafe, polluted traffic-choked artery between two major crossings and their approaches) don’t exactly make it conducive to more intensive development. Chinatown is beginning to gentrify, but that’s not where the gentrification is occurring.

      • Nathanael says:

        I think in terms of electrification, the priorities should be:
        (A) Electrify the Albany-Penn Station line with overhead, 25 kV 60 Hz
        (B) Electrify the MBTA South Side (overhead 25 kV 60 Hz).
        (C) Electrify the MBTA North Side (overhead 25 kV 60 Hz).
        (D) Figure out how to convert the rest of Metro-North to overhead, using conductor bars if necessary. Metro-North uses a nonstandard form of third rail used nowhere else in the world, which cannot be good for maintenance costs.
        (E) Convert Metra Electric and the South Shore Line to AC (overhead 25 kV 60 Hz)
        (F) Fully electrify the NJ Coast Line with overhead 25 kV 60 Hz
        (G) Convert the NY-Trenton line to 25 kv 60 Hz, allowing NJT to forget about 25 Hz.

        This gets rid of the Totally Obsolete Systems north of SEPTAland while making some critical improvements. Without trying to reform the probably-hopeless LIRR.

  2. […] 7/23: hey everyone, go read Queens Transit, both for the regional rail posts (which I don’t fully agree with, but think are thoughtful and interesting), and for the […]

Leave a Reply

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

You are commenting using your 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

%d bloggers like this: