Bergenline Avenue and New Hudson Tunnels

The main street of Hudson County from Jersey City north is Bergenline Avenue. It passes through the densest cities in the US (denser than New York, which is weighed down by outer-urban areas), and hosts frequent jitney service. Last decade, New Jersey began to document jitney service in North Jersey, producing a report in 2011 that identified major corridors; Bergenline is the busiest, with a jitney almost every minute, and almost as frequent additional jitney and New Jersey Transit service on the northern part of the route running into Manhattan via the Lincoln Tunnel. This was discussed extensively on Cap’n Transit’s blog three years ago, and I thought (and still think) Bergenline should eventually get a subway line. I bring this up because of a critical tie-in to Bergenline’s transit service: new mainline Hudson tunnels. If the new tunnels are built to host regional rather than intercity trains, then they should also make a stop at Bergenline to allow for easier transfers from the buses to Manhattan.

Unfortunately, there are no estimates of ridership on the Bergenline buses. The 2011 report did rough counts of passengers per hour passing through a single point, but that is not directly comparable to the usual metrics of ridership per day or per year. Moreover, the report assumed there are 16 passengers per jitney, where, at least in Cap’n Transit’s experience, the jitneys on Bergenline are considerably larger, in the 20-30 passenger range. Either way, they’re smaller than full-size buses, which means we can’t just compare the frequency on Bergenline with that on busy New York bus corridors. However, a bus in that size range almost every minute, both peak and off-peak, is bound to have comparable ridership to the busiest buses in New York: the single busiest, the M15, runs articulated buses every 3 minutes at the peak and every 4 off-peak.

There are several corridors heading into Manhattan. According to the summary on the report’s PDF-page 51, Bergenline has jitneys heading into Port Authority every 2-4 minutes at the peak, and New Jersey Transit buses (routes 156 and 159) every 5 minutes. Paralleling Bergenline, JFK Boulevard East has a jitney every 4-5 minutes (with larger vehicles than on Bergenline), and a New Jersey Transit bus almost every minute at the peak (route 128). There is also very frequent New Jersey Transit bus service, more than once per minute between routes 156, 159, and 166, running nonstop to Port Authority at the peak; unlike the jitneys, New Jersey Transit bus service is extremely peaky, with the combined routes 156 and 159 dropping to a bus every 15 minutes, and the Boulevard East routes (165, 166, 168) dropping to a bus every 9 minutes.

From the New Jersey Transit schedules, peak-hour buses spend 18-19 minutes getting into Port Authority from Bergenline, and 14 minutes getting into Port Authority from Boulevard East. In contrast, a train station located under Bergenline would have service to Penn Station taking about 3 minutes. Trains go through the existing older tunnel at about 100 km/h, and the new tunnel could support at least the same speed, while a through-running service plan would simplify the Penn Station interlockings enough that trains could enter and leave the station at speed. Even allowing for transfer time and for additional wait times, which are very short at the peak anyway, this represents an improvement of more than 10 minutes.

It goes without saying that the service should be frequent and affordable. The fare should be the same as on the subway, with free transfers. There’s some precedent in that PATH charges similar fares to the subway, but free transfers, a basic amenity in regions with integrated transportation planning, would be new to New York. At the peak, all trains would stop at Bergenline, since there’s not enough capacity to mix stopping and nonstop trains on the same tracks given expected traffic. But even off-peak, all trains should continue stopping at Bergenline – as well as at Secaucus – in order to maintain adequate frequency. Given how dense and close to Manhattan the area is, 10 minutes is the maximum acceptable headway, which corresponds to the combined off-peak frequency of all New Jersey Transit trains into Penn Station today.

While the busiest trunk line does not even enter Manhattan, the presence of fast, frequent regional rail with competitive fares is likely to change travel patterns. This is not the same as transit-oriented development: I am not assuming a single new building on top of the Palisades. Instead, some people who live and work in northern Hudson County would shift over time to working in New York, thanks to improved transportation links. In parallel, people working in New York would move to cheaper housing in Hudson County. In the other direction, companies that want to attract reverse commuters might locate to the area around the new station. The overall effect would integrate northern Hudson County into the core better, turning it into more of a bedroom community, like Brooklyn and Queens, while simultaneously concentrating its employment around the station. The upshot is that this station would already come equipped with a huge installed base of feeder buses, which run the route already without a connection to Manhattan. A longer-range plan to build a subway under Bergenline, from Fort Lee to Journal Square, would further integrate the entire west bank of the lower Hudson into the city core.

This tilts the best traffic plan for new tunnels away from Amtrak’s Gateway plan and back toward New Jersey Transit’s various flavors of ARC. First, it’s easier to build the station while the tunnel is excavated than to build the station in the preexisting tunnel. At the same time, whichever tunnel has the station should be the one without intercity trains: all peak trains would have to stop at the station for capacity reasons (there’s no room for bypass tracks), and this would slow down intercity trains unacceptably. Put together, this means Amtrak should stay in the old tunnels and all traffic in the new tunnels should be regional.

Second and more importantly, a high-grade new tunnel pair from New Jersey to Penn Station should also continue onward to Grand Central, with trains running through to Metro-North territory. The importance of through-running and good service to multiple urban nodes is greatest for local service and smallest for long-distance service. In Paris, the RER involves through-service for shorter-range commuter trains; the Transiliens, which terminate at the traditional terminal stations, serve farther-away suburbs. And in Tokyo, the local lines of the JR East network run through whereas the express lines either don’t or have only started doing so recently. The reason is similar to a pattern I mentioned before about airports: at long range, people only travel to the city for functions that their region lacks, and those are usually centered on the CBD, whereas at short range, people travel in all directions. The upshot of this discussion is that a Bergenline stop is likely to add many local travelers to the system, and they should get the service that’s more useful for their needs.

Of course, a good service plan will involve through-running in both the old and new tunnels. However, through-running is more valuable in the new tunnel, going to Grand Central, than in the old tunnel, going to Long Island and the Northeast Corridor. As a judgment call, I believe that through-running to Grand Central, Harlem, and the South Bronx connects to more neighborhoods than through-running to Sunnyside, Flushing, and Jamaica. It also has better subway connections, to the 4/5/6 if to nothing else, and local riders are accustomed to two-seat rides and subway connections. Finally, under a fuller regional rail plan, including service to Lower Manhattan, Grand Central has connections to Lower Manhattan and Downtown Brooklyn whereas Penn Station and Sunnyside don’t.

In contrast, Amtrak’s plan gets it exactly backward in proposing to use the Gateway tunnel for its own trains and some additional regional trains. The only advantage of this plan is that it would be possible for regional trains to maintain higher speed through the wider-diameter new tunnel (intercity trains could raise speeds more easily, since high-speed trains are pressurized to limit ear popping when they enter tunnels). But by hogging slots in the Penn Station-Grand Central tunnel, Amtrak would force many local and regional rail riders onto trains that do not serve their destination directly and do not have an easy transfer to it.

The only drawback of this plan is cost. The station would be located deep beneath the Palisades, complicating its construction. While the access shafts are not difficult – vertical bores for elevators are simply to build – the station itself would require blasting a cavern, or using a large-diameter bore. The cavern option is not cheap. I am not going to try coming up with a cost estimate, but I will note that the station caverns of Second Avenue Subway Phase 1, which are built cut-and-cover rather than blasted from inside, are around a billion dollars each. A large-diameter bore is more attractive, but is more expensive than twin small-diameter bores if there are no stations, and may well have difficulties emerging at the Manhattan end.

Without reliable estimates for either the incremental cost or the incremental ridership, I can’t say whether this is a cost-effective proposal. I suspect that it is, given the high ridership of the Bergenline buses and the high density of the region. Part of what makes an S-Bahn or RER system successful is its service to urban neighborhoods and not just suburbs and CBDs, and Bergenline could be a good addition to the system that the region should be building.

70 comments

  1. Ryan

    There’s so much pent-up demand and every one of the far-too-few Hudson River Crossings are so far over capacity that, frankly, the answer isn’t “which new rail tunnel do we build,” it’s which tunnel gets built first.

    Room must be made for bypass tracks. There’s basically one place on the entire Northeast Corridor where the case can be made that six tracks are actually necessary, and it’s across the Hudson between New York and Newark.

    • Alon Levy

      Why must room be made for bypass tracks? This is bored tunnel; the cost of four tracks is twice the cost of two, so might as well build the second track pair to somewhere other than Penn Station.

      • Ryan

        Building another track pair to somewhere other than Penn Station isn’t going to change the need for six tracks into Penn Station. Running trains into Lower Manhattan is what stops you from needing eight tracks into Penn Station.

        You can trade Tracks 5 and 6 for more cross-Hudson subway capacity, but then you don’t need to bother building a regional rail station at Bergenline because people going there could utilize the 7 and/or whichever services from east of the Hudson feed into a Bergenline Avenue subway tunnel.

      • Eric

        Wouldn’t a 4 track tunnel (2×2) be significantly less expensive than two 2-track or four 1-track tunnels?

        • Alon Levy

          If it’s one large-diameter tunnel with 4 tracks then yes, but then you can forget about building a station inside that tunnel.

  2. threestationsquare

    The Bergenline Avenue HBLR Station was built along the existing Weehawken rail tunnel, apparently by blasting/excavating a cavern. The cost was reportedly $150M in 2004. A regional rail station would need to be longer but the cost of excavating a station on an existing tunnel (at least if you can shut it down for construction) is not necessarily prohibitive. (But as you noted it likely makes more sense to use the new tunnel for local service including the Bergenline stop and the old for intercity, for the sake of Grand Central through-routing.)

    • Ryan

      Intercity trains benefit from Grand Central through-routing as well, both for the intercity rail riders who might be seeking the East Side (and frankly Grand Central’s more than large enough to accommodate some of its space being rededicated to intercity passengers and the services they require without anyone else really noticing) and for running trains from west/south of New York to Albany and points further north.

      Honestly, you could make the argument that moving Amtrak (back) into Grand Central is a better proposal than any of the permutations of building a new station (held back mostly by the fact that East Side Access is a bottomless dumpster full of money on fire) that are currently on the table, but that’s an argument for another post.

      • Nathanael

        “Alternative G” in the original “ARC” study, tunnels from Penn to the lower level of Grand Central, needed to be done a long, long time ago.

        And I agree that intercity services really ought to be concentrated entirely in GCT, which has the space for all the auxiliary facilities, and sufficiently wide platforms — not Penn, which doesn’t have the space and has narrow platforms. But unfortunately that would require so many agreements that it probably won’t happen.

          • Nathanael

            Penn sucks if you want to get somewhere other than Midtown on the West Side. GCT is definitively better than Penn for nearly all New York City destinations other than right-next-to-Penn.

            However, the benefit of having connections to Long Island, New Jersey, Connecticut and so forth in one place is probably significant.

          • threestationsquare

            Penn has two north-south subways, at least one of which has some spare capacity; GCT has one overcrowded one. The E gets you from Penn to east Midtown.

          • Adirondacker12800

            Four if you don’t mind walking a block over to 6th.

      • Alon Levy

        Intercity trains benefit from Grand Central service, but less than commuter trains, for the following reasons:

        1. Providing two separate stations is not very useful for intercity trains, but is useful for commuter trains, where it would give riders more options and also distribute the passenger loads in Manhattan better, reducing dwell times.

        2. There’s no room for bypass tracks at Harlem-125th, and this limits capacity on the track pair out of Grand Central that hosts intercity and regional trains. This is of course also true of a potential Astoria station on Metro-North Penn Station Access, but a lot of regional trains would divert to the LIRR and not have to deal with Astoria, whereas every train leaving Grand Central going north has to stop at or pass by Harlem.

        3. The Penn-GCT link itself is useful as urban rail. Both intercity and regional trains could provide it if Amtrak agrees to not check tickets between the two stations (and between South Station and Back Bay), but regional trains can offer direct links to Penn from Harlem and the South Bronx whereas intercity trains can’t. This is what I mean when I say that the link is so important for local service that every train using it should be regional rather than intercity.

        • Ryan

          No argument on #1, but I’ll point out that the argument I was touching on was to use Grand Central as the new replacement Amtrak station, moving everything except maybe a ticket window or two out of Penn Station – the same way the plan seems to be developing for “Moynihan,” or was for Penn Station South, or any of the other plans for new buildings. To use the Boston analogy, Penn would become the Back Bay equivalent and Grand Central the South Station equivalent – not the other way around.

          #2 is a problem – but it’s a solvable problem.

          Per #3, I would expect everything to move through the link at the same (relatively low) speeds and so, with no need to arrange for overtakes, it won’t matter if a limited number of slots get turned over to Amtrak. I somehow doubt that we’re ever going to live in a world where there’s more than 10 intercity tph in each direction between Penn and GCT – or if we do end up living in that world then Amtrak’s so flush with cash it won’t even notice the price tag on extending the six-track stretch through the Penn-GCT link. Until that happens, regional rail only getting a train every 72 seconds in each direction instead of every 60 isn’t really that huge of a loss, especially with the through traffic between Penn and GCT representing just half of all regional rail traffic with the other half being split between the legacy Penn-Sunnyside tracks and a new tunnel through Lower Manhattan to Brooklyn.

          • Alon Levy

            Well, you know my thoughts on Moynihan.

            Ad 2, it’s not easily solvable. Harlem has four tracks. Yes, it can be rebuilt with six, but it’s disruptive.

            Ad 3, look at worst cases, not average cases. Say the capacity of the two-track tunnel is a train every 2.5 minutes, and there’s an intercity train every 10 minutes. Then six times per peak hour there’s a five-minute gap between trains, which given that it’s basically a second 42nd Street Shuttle, that’s a problem.

          • Ryan

            It’s frankly not worth building a two-track link between Penn and GCT. If we can’t find the money four a four-track link we shouldn’t bother until the money’s available.

            The worst case scenario for a four-track link if you figure all tracks are functional and divided strictly into two-track one-direction pairs is likely 40 trains (20 * 2 tracks) per hour – once every three minutes per track and once every 90 seconds overall. If you then say that a full quarter of those slots go away for intercity trains, you’re left with ten gaps of three minutes between trains every peak hour. That’s about as good as the 42 St Shuttle today.

            Hell, even if you’re going to say the worst case is merely 15 per hour per track, you’re still getting trains every two minutes and the wait would never be longer than four minutes for the next shuttle in either direction.

          • Alon Levy

            Do people ever build four-track connections like this nowadays? The RER lines are all double-tracked. The issue is that if there’s money to build two lines, then there’s money to build them on two different alignments.

          • Ryan

            You yourself have argued for a four-track North-South Rail Link, and I somehow doubt that Boston will ever generate more through traffic than New York City. Nor will any other city in America, most likely.

            I’m going to wager that most of the places throughout the rest of the world that need a four-track connection have had it for decades now, much like Penn and GCT should have been connected since back when the original Penn Station was built.

            I’ll restate what I said at the beginning of this conversation: I’m not arguing to build four tracks here or six tracks across the Hudson because I think that’s more worthwhile than building an all-new alignment somewhere else. In fact, I’m arguing to build the all-new alignment through Lower Manhattan and Brooklyn as well because without it, you’re going to need eight tracks into Penn, not six. There’s simply that much demand to cross the Hudson, and all of its existing crossings are woefully insufficient in meeting that demand.

          • Adirondacker12800

            The solution to No. 2 is a tunnel from North White Plains to Rahway. NEC-o-the-Future and all the previous studies carefully tip toed around the capacity needed for the branches. Or the capacity needed for the trains to Chicago, Toronto, Montreal and Atlanta. Or the two extra trains per hour to Maine that will be passing through New Haven. One via Providence and one via Springfield.
            Penn Station Access for Metro North is a stop gap. In 2040 they need more Park Avenue Tunnel again. Around 2040 Gateway isn’t enough. There’s gonna be a lot of tunnels that need to be built. If Manhattan is gonna have a gazillion intercity passengers it makes sense to send 20 percent of them to Wall Street and 40 percent of them to Penn Station and 40 percent of them to Grand Central.

          • Adirondacker12800

            The bigger question is why do people want to go from Grand Central to Penn Station? Both of them will have service for Metro North, the LIRR, NJTransit and Amtrak. It’s not particularly useful if you want to get to Rockefeller Center.

          • threestationsquare

            @Alon “Do people ever build four-track connections like this nowadays?”: The Tiergarten Tunnels (completed 2005) south from Berlin Hauptbahnhof have four mainline rail tracks.

          • Alon Levy

            Also, about the Boston four-track rail link… the issue there is that it’s difficult to build it with just two tracks and connect it to every MBTA branch. At the southern end there have to be at least two portals, one pointing toward Back Bay and one toward JFK-UMass, which means that two bores are needed anyway. If the bore starts after the two lines come together, then there’s not enough room to clear all the underground infrastructure at South Station; the NSRL plan calls for 3% grades, even with multiple portals, because that’s how deep the tunnels have to go to duck under I-93.

            More generally, Boston is like Philadelphia, which also has four tracks, and unlike New York, in that there really isn’t a second alignment. New York’s always had many terminals outside Manhattan, plus Grand Central: Long Island City, Flatbush, St. George, and too many to list in Hudson County. Penn Station replaced Exchange Place and LIC, and over time most other Hudson County terminals were abandoned, but Flatbush and Hoboken still exist, and all the lines that go to Hoboken can also go to Pavonia (or for that matter Exchange Place with a short tunnel). Boston is not like that: it built two union stations early, so all that’s needed is to connect them. The only loose end is East Boston and the remains of the Eastern Railroad, and sending one of the two bores from the south there instead of to North Station is an interesting idea but I think would be far too expensive and also mismatch ridership, with roughly even ridership levels in the two bores from the south but far more ridership from North Station than from East Boston.

            EDIT: I forgot – there’s an old post of Cap’n Transit about all the railroad stub terminals of postwar Paris and New York, which was one of the influences on my thinking about regional rail. In contrast, Boston has two terminals, plus the Eastern ROW, and at least by the 1970s Philadelphia also had just two, plus some lines heading into Camden. In the same way that I ask about the three second cities of the Northeast “where’s their Utica?” or “where’s their Second Avenue?”, I can ask “where are their Pavonia and Flatbush?”. (Okay, DC has much less legacy commuter rail infrastructure, since it used to be much smaller than Boston and Philly, so it only needs catenary wires and through-routing between MARC and the VRE on existing double track, but the same issue with not really having room for multiple independent alignments is there.)

          • Ryan

            Sure, but again, the important distinction here is the overwhelming capacity requirements owing to demand generated by the many disparate services originating from as far out as eastern Long Island and southern New Jersey.

            Flatbush and Exchange Place still exist, yes: they must be connected by a new alignment underneath Lower Manhattan. That’s going to suck up about a sixth of the regional rail demand from everywhere to Manhattan in the form of people specifically trying to get to or around Wall Street. That’s not going to solve your whole problem with getting people across the Hudson. It isn’t even going to solve a third of your problem, you still need another four tracks plus the two legacy tracks that exist today.

            Long Island City still exists, yes. You could make an argument that Long Island City ought to be connected directly to Tracks 1-5 in Penn Station by a new two-track tunnel in order to hoover up most of the LIRR traffic clogging the existing East River Tunnels that isn’t going to maybe some day eventually start running into Grand Central instead. That doesn’t mean you don’t still need six tracks crossing the Hudson. It might not even mean you can get away with dropping the Penn-GCT Link to two tracks: demand is presumably about equal between the branches east out to Long Island and north upstate/to Connecticut, but even if there is clearly a majority demand originating from Queens or the legacy NEC it certainly isn’t going to be overwhelming enough to justify a 6/2 track split instead of 4/4.

            Hoboken still exists. Unfortunately, there’s no good place to send tunnel tracks from Hoboken across Manhattan and into Brooklyn, unless you’re willing to go through an awful lot of far-more-valuable subway tunnels.

            The fact is that here – maybe nowhere else in America but certainly in New York City – it isn’t unreasonable to bet on and plan for a future where the peak hour TPH is 80 each way between New Jersey and Long Island and 60 more heading upstate or into New England. There’s too much demand concentrated on a relatively small area and not enough alternate alignments – somebody’s going to need to get bypass tracks or express tracks or whatever else you want to call additional tracks. The 2-track RER setup is just not sufficient for moving the sheer volume of traffic that New York City’s going to see in 2040. There’s no way to divide this so that we end up with three neat and tidy pairs of tracks on three separate east-west alignments.

            There’s an argument to be made, perhaps, that you might want to stay with two tracks between Penn and GCT and four tracks across the Hudson into Penn and expand the Atlantic branch to four tracks to feed a four-track tunnel through Lower Manhattan instead. I think expanding the Atlantic Branch is more trouble than it’s worth, but that’s the alternative.

          • Adirondacker12800

            You could make an argument that Long Island City ought to be connected directly to Tracks 1-5 in Penn Station by a new two-track tunnel in order to hoover up most of the LIRR traffic clogging the existing East River Tunnels that isn’t going to maybe some day eventually start running into Grand Central instead

            Tracks 1-4. according to what I’ve read on railroad.net, are forever stub tracks. The Sixth Avenue Subway is in the way. And they are being used by those pesky New Jerseyans. And the occasional Empire Service train. As are the track that will be built for Gateway. One of the other rumors floating around is that the LIRR is going to abandon service to Long Island City.

            You can suck up lots of the traffic going to Penn Station or someday Grand Central by letting the people who are clogging the subway to get to Wall Street by sending LIRR trains to Wall Street. And the people clogging PATH in New Jersey by sending their trains to Wall Street.

            If I remember correctly the LIRR is proposing a 60/40 split between Penn Station and Grand Central, with increased service. There’s 6 tracks through Woodside. Trying to stuff 10 pounds of train through that so 13 pounds of tunnel is filled won’t work out. Metro North is drooling to suck up any freed up capacity. And I’m sure Amtrak is eyeing it. The improvements in Connecticut, Western Massachusetts and Vermont will be completed by the time East Side Access opens.

            Hoboken still exists. Unfortunately, there’s no good place to send tunnel tracks from Hoboken across Manhattan

            No there isn’t. It’s a ferry terminal and some of the tracks get flooded in very high tides. Tear down the suburban mall at Newport and connect it to the existing lines just a short distance away, There is plenty of room in the cut the PATH trains use and the cut the Erie built is unused. Every line in New Jersey would connect to Wall Street. And everything on Long Island except the Port Washington Branch. 20,000 people an hour at peak that use Wall Street are 20,000 people that aren’t in Penn Station. Or in the other tunnels.

            Upstate has access to Manhattan via the Empire Connection. Someday far in the future they could run some trains to Grand Central and most trains to Penn Station. That tunnel from North White Plains to Rahway looks really good to me because with a stop at Grand Central and stop at Penn Station the 50-ish miles takes half an hour. And I’d probably have through service to Washington DC and maybe Harrisburg. and through service to Cleveland, Toronto, Montreal and Boston. Unfortunately I’ll probably be dead in 2040.

          • Alon Levy

            The plan for Alt G was to connect tracks 1-4 to Grand Central, not to build new tracks for the Grand Central connection.

          • Alon Levy

            We don’t actually know why it was rejected – Stephen Smith’s been trying to FOIA that for years. But it couldn’t have been about that, because Alt S, which included a new tunnel pair across the East River, again starting from tracks 1-4, was not rejected initially.

          • Adirondacker12800

            Get while you can. Somebody else has started to use arctunnel.com and eventually the older stuff from the NJTransit site will fall off as too old.

            Just because foamers get all frothy at the prospect doesn’t make it so.

            http://web.archive.org/web/20101025135306/http://arctunnel.com/pdf/library/ARC_MIS_Summary_Report.pdf

            Not enough capacity:

            Page 25 of the file or page 19 by the report’s page numbering

            “The results yielded a conceptual service plan that indicated an increase of 13 additional NJ TRANSIT inbound trains (Secaucus to Penn Station) over the Post-Secaucus Service Plan during the AM peak hour, for an inbound total of 36 trains (34 NJ TRANSIT, two Amtrak) occupying 38 slots. Additional service beyond the 13 would make the entire operation unreliable.”

            Not enough capacity, costs too much and too risky:

            Page 40 of the file or page 34 by the reports’s page numbering.

            “Although the conversion of Metro-North and NJTRANSIT operations from stub-end to flow-through movements was determined to be physically feasible, there would be impacts on NYCT subway structures and the operations support systems at Grand Central Terminal. Uncertainties over the extent to which these impacts could be mitigated could not be resolved satisfactorily during the Phase 3 effort. Alternative G would offer the smallest incremental increase in trans-Hudson train capacity among the alternatives and would create complex train operations that could affect the operational reliability of the respective raiAlthough the conversion of Metro-North and NJTRANSIT operations from stub-end to flow-through movements was determined to be physically feasible, there would be impacts on NYCT subway structures and the operations support systems at Grand Central Terminal. Uncertainties over the extent to which these impacts could be mitigated could not be resolved satisfactorily during the Phase 3 effort. Alternative G would offer the smallest incremental increase in trans-Hudson train capacity among the alternatives and would create complex train operations that could affect the operational reliability of the respective railroads. Construction of Alternative G would require negotiation of easements or purchase of a large number of Manhattan properties.
            In addition, the physical impacts of construction on the ability to maintain existing operations appear to be significanlroads. Construction of Alternative G would require negotiation of easements or purchase of a large number of Manhattan properties. In addition, the physical impacts of construction on the ability to maintain existing operations appear to be significant.

            Somewhere in there is a nice little blurb about wreaking havoc on the Lexington Avenue subway. Just because Grand Central was built so that tracks could head south doesn’t mean you’d want to do that today. Or that the NYCTA didn’t and put the express tracks for the 6th Avenue line where the PRR was plotting to put more tracks if they needed them.

            It’s not a viable option.

            NEC-o-the-Future is more or less Alternative-P. With the dirty little secret that they need a tunnel all the way to Rahway not mentioned.

          • Joey

            The capacity analysis assumed no through running though – NJT trains terminate at GCT and MNR trains terminate at Penn. Through running increases capacity in this scenario for two reasons – (1) You don’t have two separate services competing for space in the two-track connection (e.g. the slot that a MNR train takes up from Penn to GCT could be occupied all the way from New Jersey and (2) Turnbacks invariably create conflicts in station throats unless you have flying junctions.

          • Adirondacker12800

            It assumed through running for at least some of the trains.

          • Joey

            Through running should allow the new tunnels to run at least 24 tph in each direction. The two sets of tunnels together should be able to push 48 tph in each direction with full through running.

          • Adirondacker12800

            NJTransit was proposing 26 for ARC and Amtrak is promising 25 for Gateway.

          • Joey

            Okay, but that only strengthens my point – if they think they can only reliably get an additional 13 tph out of the Penn-GCT connection, then there’s something very off about their assumptions.

          • Ryan

            Nah, I don’t think there’s anything off about their assumptions.

            A lot of things are off with regards to their organizational incompetence but I’m not sure today’s MTA can be reliably counted on to run 20 tph per track, never mind 24 and certainly not the 27~30 that represents the ceiling of what’s possible with modern signaling and switching technologies.

            But just like it’s a safe bet that we ought to be planning to meet 60~80 tph of demand in every direction, it’s a safe bet that the relevant authorities are either going to be forced to correct their own incompetencies or be forced out in favor of groups that can. Whether that happens before or after we cross the 40 tph demand line is an open question, but it’s not if it’s when.

          • Alon Levy

            The MTA runs 24 tph on the subway.

            Anyway, a second tunnel to Penn Station and Grand Central, and an Erie-Fulton-Flatbush tunnel, together with the old tunnel, give you that 60-80 tph capacity.

          • Adirondacker12800

            The usual number tossed around for Penn Station is 23 an hour. There are claims,that in December when NJTransit adds a few trains, they manage 26 in one 60 minute period. The ARC studies claimed 48 or 52.

          • michael.r.james

            Penn Station to Grand Central?

            Instead of multi-billion dollar bored-tunnels etc. how about thinking laterally a bit: these are ≈1.9 km apart (along 34th, up Park Ave.) which is longish but entirely feasible by travellator. Obviously a series of them. Make it a city feature by putting them above the street, in a perspex bubble. Have 3 “tracks” and add gardens–like the High Line/Promenade Plantée–and cafes etc. Not only would it be vastly cheaper but it would allow entry/exit at a whole series of points along this busy route.

            I can’t find what the longest travellators in the world are, probably at some airport, but the two that link (underground) Chatelet to Forum Les Halles span about 450m. The Hong Kong Central–Mid-Levels escalator is possibly the most comparable concept; it spans 800m in a series of standard escalators (open-air with a domed polycarbonate roof).

          • Alon Levy

            1.9 km is too long for a travellator. Even 500 meters elicits groans from people who have to make such transfers regularly.

          • Adirondacker12800

            a moving sidewalk is too slow if you want to get from Boston to DC in three hours. It’s too slow if you want to go from Jersey City to 47th and Third to go to work.

          • johndmuller

            Focusing primarily on the mid-town aspects of this (not to disrespect the Brooklyn-Wall Street-Jersey routes), it seems likely that at least some tunnel(s) will get built under the Hudson roughly along the current route to augment the pair already in place to NYP. The barebones plan so far seems to be to create more stub-end tracks at the southern side of NYP. If Amtrak were, as it currently appears to be, the main sponsor of this construction, it could certainly expect to control plenty of the new capacity, only the glibbest of spin doctors could paint this as a big improvement to through-NYC traffic on the NEC.

            To get any boost to through traffic, they would need to extend some of those new tracks across Manhattan under 31st Street; another switch tree such as the ones leading to the 32nd and 33rd Street tunnels could feed the new tracks and some of the existing ones into a new tunnel under 31st St. I can think of three obvious destinations where such tracks could be going – a) straight across Manhattan and under the East River to Sunnyside; b) part of the way across, then north to GCT (original and/or new levels); and c) part across and then south to lower Manhattan to connect to whatever ultimately happens with the Long Island-Wall Street-Jersey City/Hoboken thing.

            It would be surprising if all three of those things happened, but all three should at least be thought about at this stage so that whatever designs are considered are compatible with all three destinations. For example, you’d probably want to have more than two tracks, thus perhaps two or more levels of tracks or two or more tunnel bores in order to allow for this to happen.

            In any case, at least some of the longer term possibilities should remain in mind before the tunnel suddenly becomes a done deal a la the Tappan Zee replacement and all loose ends get swept away in some political frenzy. Remember both Cuomo and Christie consider themselves to be upwardly mobile and could get together without warning to do something big at a politically poignant moment. If they suddenly ‘found’ a billion or two of the Port Authority’s money, they might be able to steamroll something through before you could say ‘7 train to Secaucus’.

          • Adirondacker12800

            Christe had 3 billion of the Port Authority’s money, 3 billion from the Federal Government and a bit over a billion from the Turnpike Authority. He pissed away the 600 million that had been spent of studies and engineering. It was scheduled to be completed in 2017. Amtrak might be thinking about a draft environmental report by then.

            If the NJTransit trains are being shunted off to the new tracks at the stub end terminal at the eastern end of the new tunnel that leaves scads of space in the center of the station for Amtrak to use for services that go through Manhattan.

            Wall Street is the country’s third largest employment center. Bore tunnels from Jersey City to Brooklyn and it would be one of the busiest commuter rail corridors in the country.

          • Nathanael

            Alon: a *third* 42nd Street Shuttle. It’s OK to have a five-minute gap between trains. Also, you can definitely run enough trains through Harlem on four tracks. Four tracks is ridiculously huge capacity if you separate locals and expresses

            The bottleneck in an “everything through GCT” plan is more likely to be the at-grade junction where the Harlem Line diverges from the Hudson line, with inbound Harlem and outbound Hudson trains interefering with each other, and having to be re-sorted into locals and expresses at the same location. The at-grade terminal approaches to GCT with every outgoing train crossing every incoming train could also be the bottleneck.

          • Nathanael

            Oh, to unpack my statement about a third 42nd St. shuttle, I am of course referring to the #7 subway.

          • Nathanael

            For reference, the official stated reason for rejecting Alt G, in the original “alternatives analysis”, was that the land it would go under was the most expensive in the US and they didn’t want to try to deal with the property owners. That’s what they said. I am inclined to believe that that was their reason.

      • Ed

        Unlike many people on these forums, I think making the old central post office into the main entry point to the Penn Station Amtrak tracks was a good idea. Intercity rail passengers should arrive from and leave a city in somewhat more style than through a basement. As it is, the Amtrak part of the existing Penn Station is somewhat less awful than the commuter rail part. And the existing Penn Station is cramped and some rearrangement would have to be done anyway.

        But making Grand Central the intercity rail hub for New York is a much better idea, it has the style, its a much better location to arrive into, and it resolves the capacity problems of Penn Station much better, and especially given the money poured into East Side Access I’ve always wondered why this wasn’t done.

        • Adirondacker12800

          Amtrak is aiming for 4 times as many passengers. Without including the demand that would come with better service to Springfield, Albany and beyond. Or Harrisburg and beyond. Or Alexandria and beyond. Send half of them to Grand Central and there will “only”: be twice as many in Penn Station as there are today.

          • Nathanael

            If you make Grand Central *the* stop (like South Station in Chicago), leaving Penn in the position of Back Bay, the vast majority of passengers will go to GCT.

            More passengers are heading to the East Side than the West Side *anyway*.

          • Nathanael

            Grrr, “like South Station in Boston”. Anyway, the actual studies, last I checked, said that more than half of passengers would choose Grand Central given the choice between GCT and Penn. This is *even more true* of intercity passengers than of passengers from the region.

          • Joey

            The problem with using GCT for intercity trains is that the Park Ave tunnels don’t really have any capacity for them. You could free up a bit with Penn access for the Hudson and New Haven lines, and by grade separating the junction with the New Haven line, but it’s not going to be enough to move *all* intercity operations to Grand Central.

          • Joey

            The last time I checked, Metro North ran 56 tph in the inbound peak direction, more than two tracks can handle (which is why they do 3+1 at peak hours). That means that the third peak track does have some spare capacity, but then you can’t segregate locals and expresses in the reverse-peak direction. Which is a problem because all reverse-peak trains need to stop at 125th St.

          • Adirondacker12800

            If you are on an Amtrak train that passes through Manhattan and you are coming from or headed to New Jersey or Connecticut it’s much easier and faster to just use an Amtrak station in New Jersey or Connecticut, And changing to NJTransit or Metro North involves walking across the platform or at worst going up and over/down and over.
            There’s six and half million people in the outer boroughs. Grand Central has access to the IRT, the IRT and the IRT. Penn Station has access to the IRT, IND and BMT. Getting to World Trade Center or Rockefeller Center is much easier from Penn Station than it is from Grand Central. And Long Island.

            Amtrak doesn’t want to use the Park Avenue Tunnel. Amtrak wants to use the tunnel to North White Plains. With a deep cavern station under Grand Central and deep cavern station under Penn Station.

          • johndmuller

            This kind of thing is why Union stations got built; if you move the intercity trains to GCT, then people connected to NYP are inconvenienced instead of the current setup where people connected to GCT are taking taxis and subways between stations. Of course it should be built in Union Square, more or less equidistant from GCT, NYP and the yet to be built Wall Street superstation, tying in LIRR/Brooklyn and NJT/JC-Hoboken (not to mention the SIR). Then we add in the Bergenline subways [north and south branches] which will be connected to the L line extension, heading conveniently straight to the NYC Union Station at Union Square (Union-Union for short).

          • Eric

            Sorry, but nobody’s going to build a brand new central rail station at or under Union Square, when Penn Station already has 21 tracks and GCT has 67 tracks, and both have large passenger waiting areas, and Union Square has nothing except for of expensive existing buildings that would have to be torn down.

          • johndmuller

            Yes Eric, Union-Union is unlikely; that was sort of my point.

            Nonetheless, a station has been proposed more or less there as part of a thru-running scheme I believe, perhaps even by Alon, whereby Metro North would break out of GCT to go down to Wall Street and then to Staten Island. Of course, there would need to be a Wall Street Station (and naturally a station on SI). Of course, the Wall St. Station would be the junction of a line connecting the Atlantic Terminal to yes, another new station in Jersey City (because Hoboken Terminal isn’t suitable apparently). So how many new stations is that — I count 4 — not to mention the extensive work at GCT and Atlantic. It goes without saying that we are going to mess with NYP anyway, so what are we missing?

            Well obviously, we are missing the money – I’m just talking about what sort of amounts to a minimal network that would connect most of what needs to be connected to have a smooth running system. It still leaves things to be desired in terms of getting out of Manhattan to the northeast, i.e to Connecticut and the rest of New England, and it doesn’t get any freight across the Hudson either,

            I my last couple of posts here I talked about running tunnels east on 31st St. from post Gateway NYP to go over to Sunnyside, but also providing design consideration and/or portals toward GCT and also to the south. Adding another southbound tube from GCT to join the southern branch from NYP would be pretty much all the Manhattan trackwork needed.

            You could do worse than put a station at Union Square, though, and not just because of its subway connections. For one thing, it is a foot in the door for the possibilities of the Wall Street/Atlantic Ave/Jersey city/Staten Island thing, without all the astronomical costs. It might not itself be all that expensive assuming it could be done as a cut and cover project under the park – given enough talking-nice with the park people.

            Yes, all this stuff is ridiculously expensive; even one of these projects seems too expensive sometimes. All the same, we have to do at least some of these projects.

          • Alon Levy

            Union Square is fine as a regional station – think Etoile or Nation or Auber. It’s not terribly useful as an intercity station. It’s in a secondary CBD, and there’s really no space for more than one platform track per tunnel track, whereas for intercity service it’s better to have two at the central station. The connections to Jersey are also kind of a disaster.

  3. Patrice

    Good article about the sheer density of population in New Jersey. I doubt that the state DOT has studied jitney use much. I disagree about the Gateway vs ARC assessment though. The 2 Hudson tunnels are simply too old and Amtrak needs to run more trains now, let alone shutting 1 tunnel to do extensive rebuilding (unacceptably slashing capacity). NJ Transit also has unmet capacity needs. So that means 4 new tunnels which would be expensive, but given the enormous population, importance and density of NY metro area, they would not be a wasted investment.

    • Alon Levy

      Both ARC and Gateway are two-track projects. The Hudson tunnels are not too old to run trains; they require more maintenance than new tunnels, even post-Sandy, but are structurally sound.

  4. anonymouse

    A regional rail stop under the Palisades definitely sounds like a very useful idea, but I’m not sure that it would be that great for getting huge numbers of people to Manhattan. The regional trains are already likely to be full (or else why run them?) by the time they get to a Bergenline station, and there won’t be much room for those riders to cram on. And if there is room, that means you’re hauling air for 50 miles just to make sure people can ride the last 1 mile, which doesn’t seem very efficient. That said, it’s still a useful link to provide access to the rest of New Jersey and other regional destinations (JFK? The Bronx? Stamford, even?)

      • Adirondacker12800

        If they are riding the subway to the commuter rail station there’s enough of them to just run the subway all the way into Manhattan.
        The Carnarsie Line to someplace in “Midtown” Hoboken and then up to the stuff on top of the cliff?

        • Alon Levy

          Think of Bergenline’s future as a non-sucky G. If we’re already discussing 7 and L extensions then they shouldn’t take over the line, QB-style, but intersect it at transfer stations and keep going to Secaucus or wherever the hell there’s demand for all these tunnels.

          In fact, the reason I came up with the commuter rail station idea is that I was considering 7 extensions to Jersey, realized that 7+Bergenline would be an effective feeder, and thought of ways to recreate that without a 7 extension.

          • Adirondacker12800

            As Anomymouse pointed out the reason they want more tunnel into Manhattan is that the trains are already standing room only. There isn’t enough room on them to dump a whole subway train full of people onto them. If you aren’t dumping whole trainloads of people onto them then you don’t really need to build a subway.
            How ’bout stop dumping suburbanites onto PATH, just let them transfer between trains to Wall Street and trains to Midtown out in the suburbs? Then reconfigure Hoboken so that there is a midtown stop on PATH and then it runs up Bergenline? If they want to go out to the suburbs they change at Newport. Tearing down a suburban style mall is a lot cheaper than carving caverns in the Palisades. And gives everyone a one seat ride to any PATH station except Newark and Harrison. Well they’d have a one seat ride to Newark on the suburban trains… probably one SRO ride to Newark…

  5. Pingback: Today’s Headlines | Streetsblog New York City
  6. Zack Rules

    Neat thoughts! What type of tunnel is the Amtrak Gateway? Bored or immersed tube? If it is immersed tube, why couldn’t a pair of subway tracks be built into the tube below the heavy rail tracks, ala East Side Access? Or maybe the PATH train could finally be absorbed by MTA subway and then the Hoboken spur could extend to Bergenline?

    • Alon Levy

      I don’t know. The ARC plan was bored rather than immersed, and I presume Gateway will be the same; the only way to get across the Palisades is via bored tunnel, so immersed is unlikely.

      Bergenline can connect to PATH even without any PATH/NYCT integration, but I’m somewhat skeptical about it. For one, it’s a detour from Journal Square, which is a bigger bus hub. Hoboken has a lot of rail ridership, but remember that even if the only tunnel that’s built is ARC/Gateway, Hoboken will still lose a lot of ridership to the new tunnel.

  7. ted

    If interested, October 2014 median weekday one way passengers by route:

    128: 6012
    156: 6177
    159: 10241
    165: 12129
    166: 14402
    168: 2999

  8. Pingback: The Sixth Borough Subway | Itinerant Urbanist
  9. Pingback: RRNY – The Core Stations – urbanramblesblog

Leave a Reply

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s