New York Can Save Billions on Rail Tunnels If LIRR Agrees to Share Track

RPA proposes much-needed integration of regional rail in its Fourth Regional Plan, but doesn't push for the politically challenging reforms that will be necessary to build it efficiently.

RPA's Fourth Regional Plan calls for a total of eight regional rail tracks under the East River (adding two to the four existing tunnels, plus the two East Side Access tracks under construction), but comparable service could be provided with six tracks if LIRR shares its infrastructure. Map: RPA
RPA's Fourth Regional Plan calls for a total of eight regional rail tracks under the East River (adding two to the four existing tunnels, plus the two East Side Access tracks under construction), but comparable service could be provided with six tracks if LIRR shares its infrastructure. Map: RPA

German transit planners have a saying: organization before electronics before concrete. The idea is that to make the most efficient use of resources, decisions should flow from a particular logic.

Better coordination between different agencies enables transit system design that can serve the most people most efficiently, saving huge sums in the long run. Better signaling and trains get more value out of existing infrastructure, while coming at some cost. And tunnels can expand capacity and reach, but are the most expensive type of investment.

Unfortunately, New York has never mastered the most important part — organization. The transit agencies that operate across state borders are notoriously poor at coordination. And even within the MTA, turf wars between different divisions are legendary.

So when the region does get around to building train tunnels, they are wildly expensive but still don’t enable the service improvements we should expect from new transit infrastructure.

In its recently-released Fourth Regional Plan, the Regional Plan Association proposes a regional rail network that calls for integrated operations instead of the current silos between NJ Transit, Metro-North, and the Long Island Railroad.

Integration is very much worth pursuing. There are tell-tale signs, however, that RPA didn’t put “organization” first. Namely, the plan calls for expensive tunnels that wouldn’t be necessary if LIRR agreed to reasonable arrangements to share its tracks with other services.

The RPA proposal begins with the Gateway tunnel, adding two rail tracks under the Hudson to double capacity between New York and New Jersey. That much is necessary, but it doesn’t stop there.

RPA also calls for two new rail tracks under the East River, on top of the existing four, plus two tracks that will open up by the mid-2020s, when the East Side Access project adds a direct LIRR connection to Grand Central.

These seventh and eighth tracks, connecting Penn Station to Long Island, are estimated to cost $7 billion. But there is no capacity crunch between Manhattan and Long Island that justifies more tracks, especially after East Side Access opens.

Even today, total inbound traffic at the peak is 36-37 trains per hour on the LIRR and another four on Amtrak, against a rated capacity of 48 for a four-track tunnel.

A major hurdle is LIRR’s unreasonable demand to keep all of its 37 rush hour slots into Penn Station even after East Side Access opens. The LIRR projects to send 24 trains per hour into Grand Central. The LIRR and Long Island politicians insist that future growth in demand induced by the opening of East Side Access will likely fill those 24 hourly trains to Grand Central plus 37 to Penn Station.

This is why both the railroad and Long Island politicians oppose Metro-North’s Penn Station Access project: They worry it could take Penn Station slots they think belong to the LIRR by right.

However, significant ridership growth on the LIRR is unlikely, because Long Island is not adding any housing. Nassau and Suffolk Counties both permit much less than one housing unit per 1,000 residents per year, according to HUD. Without massive transit-oriented development on Long Island, to the tune of tens of thousands of annual housing units every year, the future growth argument holds no water.

Meanwhile, a more useful piece of tunnel is absent from the plan: a connection between Penn Station and Grand Central. This connection was proposed in one of the variations for the predecessor to the Gateway project, known as ARC Alt G, and it remains feasible today. Building it would enable through-running between the New Jersey Transit and Metro-North networks.

A rail connection between Penn Station and Grand Central . Map: Alon Levy
A rail connection between Penn Station and Grand Central would enable through-running regional rail service with six tracks under the East River, forgoing the extra two proposed by RPA. Map: Alon Levy

Earlier this year, I spoke to RPA President Tom Wright about a preliminary version of the regional rail proposal. At the time, he said RPA was still considering the Penn-Grand Central tunnel for inclusion in the Fourth Regional Plan, but that the preliminary report only focused on infrastructure required for the Gateway project.

But the Penn-Grand Central link is not in the final plan, and Wright’s explanation also raises the question of why RPA included the seventh and eighth East River tracks in its earlier plan. The answer must involve the toxic relationships between the different railroad bureaucracies, and the difficult politics of getting the LIRR to share track.

If the LIRR assesses its future service plan realistically, New York can have an integrated regional rail network at far lower cost, forgoing a $7 billion tunnel under the East River. (The connection between Penn Station and Grand Central would not be free, but it’s entirely underground rather than underwater, which should make it much cheaper than tunneling under the East River.)

The politics of compelling the LIRR to share track may be challenging, but the same could be said for many components of RPA’s plan.

At the very least, someone needs to impress on Governor Cuomo the importance of integrating transit service and infrastructure. If New York is going to build a good regional rail network in our lifetimes, political heavyweights will need to insist on organization before concrete.

  • Newtonmarunner

    The maps also show that the cost of doing the additional East River Tunnels to Penn rather than the Penn-Grand Central connection isn’t just the incremental cost of tunneling under the East River but also the transportation benefits of the Penn-Grand Central connection over Penn Station South. The Penn-Grand Central connection expands the grid, and increases the number of paths to major job clusters, and relieves the subway lines by an order of magnitude more than the RPA proposal. With a Penn-Grand Central connection NJ Transit commuters have at worst a cross-platform transfer to Grand Central and its Lexington Ave. Lines (both local and express) for jobs on the Upper East Side and in Union Sq. Further, Harlem Line commuters have at worst a cross-platform transfer to Penn and its West Side Lines (both local and express) for jobs by Columbus Circle, Upper West Side, and W. 4th St. Further, those on the Lexington Ave. Lines — both local and express — can transfer at Grand Central to jobs on the commuter rail lines that go there (e.g., Jamaica Transit Center), and those on the West Side Lines — both local and express—can transfer to the commuter rail lines going through Penn (e.g., Newark, Jamaica). The RPA’s proposal only connects to the local lines, and therefore doesn’t increase capacity, reduce commute times, and make travel easier by remotely as much as the Penn-Grand Central proposal. This also should be considered in the cost of RPA’s proposal as compared to the Penn-Grand Central Regional Rail connection.

  • Stephen J. Marmon

    The Penn-Grand Central connection is a great idea but incredibly difficult and expensive to implement. The tracks are on different levels and the location of the new East Side access tracks, as well as the existing Lexington Avenue subway, would make the connection, even though it only would be less than 20 blocks long, outrageously costly.

  • Newtonmarunner

    You don’t have to hook Penn Station tracks into ESA. Alon’s 3-Line Regional Rail map doesn’t do that.

  • Stephen J. Marmon

    But you can’t link Penn w. GC on the two existing levels due to Lex line.

  • Vooch

    indeed – it’s a great & simple plan

  • Newtonmarunner

    The ARC Major Investment Study says otherwise:

  • Stephen J. Marmon

    No, I know that document. It was drawn up in 2002, long before ESA track placements were finalized. The placement of the new ESA lines make it extremely expensive to connect to Penn.

  • AC Tesla

    For the life of me, I can’t fathom why the new Gateway tunnel should cost 25 billion dollars. I live in Seattle where we are just completing a 4 lane 2 mile tunnel under the city and the tunnel bid was for about 1.2 billion dollars. Granted the TBM broke and was repaired at a cost of about $500 million more but even then that is less than a tenth of the cost of the Gateway project. No trains anywhere in the US carry more passengers than New York so I want NY/NJ To have the best transit system in the world. I just think the money is spent inefficiently. Solve that problem for the riders, the States, the taxpayers and the country. Please!

  • Larry Littlefield

    Try this.

    If NY-NJ will be charged mafia prices for a new tunnel, build it somewhere else and sink it in the river, as was done for the 63rd Street tunnel that East Side Access will go through.

  • AC Tesla

    Submerged tunnels can be built almost anywhere and floated and sunk to their final position. But whether one should be used in the Hudson as opposed to a TBM depends on the depth of the river. I would imagine there are large ships that navigate the Hudson. So my question is that an issue for shipping?

  • Liam

    This is ridiculous.

    First Metro-North refuses to share any of its existing tracks in GCT (the most of any passenger terminal *in the world*) with LIRR.

    Now LIRR is trying to block Penn Access? They’re part of the same fucking agency!! Where else in the world would you ever see infighting like this between two public agencies part of the same authority?

    Only in NY ?

  • Larry Littlefield

    Virtually no large ships sail up the Hudson. The large ships dock in New Jersey, below Manhattan. Just ferries and small watercraft, and the occasional tanker going to Albany.

    TBMs would still require lots of local labor. Moreover, unless they are going to go straight down and straight up, they would be going through loose wet soil and not rock, which would mean pressurizing the air, which would mean time in and out of depressurization.

    TBMs could be used to tunnel under the Palisades before getting to the river.

  • Larry Littlefield

    Funny how bureaucrats can screw things up and their political masters can’t get together and stop it.

    And their political masters can screw things up and the bureaucrats can’t stop it.

  • van_vlissingen

    Didn’t RPA also propose another two tracks linking Brooklyn and 3rd Ave / MetroNorth extension.

  • Newtonmarunner

    And still worth the cost — Penn to GCT compared to the RPA idea even if you stipulate to your cost assumptions. The RPA plan does not coordinate well with the subway lines (GCT has 3 trunk lines; 3rd Ave has the Carnarsie on 14th and Houston has the 7th Ave. Local on Varick) and the jobs map [3rd isn’t as close to the Midtown job cluster (particularly 6th to 8th Aves.) or Union Sq. as are GCT, Union Sq., and W. 4th]. The capacity increase, increased connections, and reduced travel time from the RPA plan is abysmal compared to connecting GCT to Penn, Union Sq., W. 4th, etc.

  • mfs

    There is a huge dissonance here in the RPA’s thinking. They call for two major governance changes — regional through-running rail and blowing up NYCT’s capital division — but assume we can’t get LIRR & MNR to cooperate on slots? If they are rebuilding Penn in the plan, why not go all the way? Such a weird combo of pulling punches and being ivory tower at the same time.

  • Stephen J. Marmon

    I agree. As I said to start off all this, “The Penn-Grand Central connection is a great idea but incredibly difficult and expensive to implement.” I’m in favor of it and the plans for through running and rebuilding Penn Station too. Just was noting that it will be very costly.

  • Larry Littlefield

    How much should New York City pay for that connection, assuming the suburbs won’t pay a dime? And should it be in the form of a cut in transit service, a cut in other services, probably for the poor, or another tax increase?

  • Newtonmarunner

    Because of the additional two underground tracks from Grand Central to Woodlawn plus the additional stations on Madison, 42nd & 3rd, 31st & 3rd, the RPA plan will still cost around the same as if not more than Alon’s 7-Line Regional Rail plan, and will give the suburbs an order of magnitude fewer connections to major and secondary job clusters and an order of magnitude fewer local and express subway line connections. Ridership won’t be remotely as high under the RPA plan than under Alon’s.

    You want savings? Eliminate skip-stop service because it’s an abomination. …

  • neroden

    LIRR’s refusal to play nice is infamous. Their unions have also been caught in multiple outright frauds. Really, *everyone* at LIRR should be fired and they should start fresh.

    Unfortunately, it’s Long Island, where corruption is the norm and politicians *openly back* corruption. You might want to look up some of the articles on local government graft in Long Island.

    It’s a cultural problem with Long Island.

  • neroden

    Basically, the construction firms are mobbed-up and overcharge as “protection money”. We should really just bring in an outside construction firm to dig the tunnel, but they’d have to bring their own heavily armed security guards.

  • neroden

    It’s still straightforward. You have to move one of the Lex line tunnels (only 1 of the four), and break through the Oyster Bar, but it’s perfectly straightforward. The ESA lines are much deeper and not relevant.

  • Newtonmarunner

    Yes, the RPA proposed a superfluous 3rd Ave. regional rail line.

    My view is to leave the Bronx alone for now. Give the Bronx service by not skipping commuter rail stops. Untangle the lettered lines first. Eventually, give the UES 8 subway tracks — including the Broadway Express Line. Then untangle the numbered lines. Then extend the Broadway Express subway Line up 3rd Ave. eventually taking over Dyre Ave. But that’s a loooong time away before the Bronx needs that kind of capacity. Don’t start with an additional two tracks in the Bronx when there hasn’t been shown a capacity problem there.

  • Newtonmarunner

    For what it does, Gateway, to be considered a good investment, should cost $3-4 billion. To be considered a marginal investment — $7-8 billion. I agree $25 billion is way, way too much and I cannot for the life of me figure out why the cost is that much.

  • AC Tesla

    The loose wet sand poses a different issue than rock but it can be addressed with a different kind of TBM such as a mixed slurry or Earth Pressure Balance Machine. Still, as long as depth isn’t a major issue a submerged tunnel probably makes more sense.. We also just built a new pontoon bridge for a major highway that runs across Lake Washington. In a way, the concrete pontoons are just like tunnel segments except they are not sunk. The massive concrete pontoons were built in Aberdeen Washington and then floated 350 miles to Seattle.

  • AC Tesla

    I was under the impression that the Mob wasn’t a problem in New York the way it use to be. So my perception is wrong?

  • Larry Littlefield

    That YouTube video shows the 63rd Street tunnel being built in 1971. IF East Side Access opens in 2021, it will have been 50 years before its lower level was put to use. That’s 50 years!

    How reprehensible is that?

    What’s the useful life on that tunnel? Do we need to start planning to build it’s replacement before it’s used?

  • AC Tesla

    Amazing…But that is nothing if you compare this project to the absurdity of the Second Avenue Subway i cant help compare New York (unfavorably) to London that is just about finished with the new Crossrail Line on time and on budget.

  • Newtonmarunner

    Even if SAS was done entirely cut-and-cover, and had shallow, bare-bones stations, it still would have costed around $1.5 billion (for 200K riders) for Phase I and probably $6-$8 billion for the entire thing. Nothing in NYC is cheap just because of the building foundations. But still way, way too much project bloat on SAS.

  • AC Tesla

    Oh no. Cut and cover would have been more expensive and disruptive. There would have been grotesque overruns from moving services. This is the lesson learned from Boston’s Big Dig. And you are wrong about building foundations. Much of Manhattan island is some of the hardest rock anywhere. The foundations of most of your skyscrapers aren’t very deep because they don’t have to be. They just bolt themselves to the rock. Where as here in Seattle which is built on 500 feet deep glacial till we have to dig deep so ours don’t tip over.

    The MTA should learn a lesson from Spain. They built their Subway tunnels with midsize TBMs which could accommodate 2 rail lines and then expanded the areas where the stations would be. I get that NYC is naturally more expensive to build than other cities but there is no excuse for it to be 10 times more expensive than Seattle or London. Neither which are cheap compared to the rest of the world. The Trade Center PATH Station was a grotesque waste of money that would have been better spent on switching or other lines. The same goes for the Fulton Station although it isn’t as bad as the PATH Station. I see Taj Mahal stations when what is needed is more tunnels and rails for more service.

  • Larry Littlefield

    “Oh no. Cut and cover would have been more expensive and disruptive. ”

    I’ve come to conclude that William Barclay Parsons, the chief engineer of the original IRT was right. What you gain in tunnel construction from deep bore tunneling, you more than lose in stations that have to bring people so far to the surface.

  • AC Tesla

    You and Parsons would be wrong today as opposed to when he said it in 1904. Deep bore subways in Manhattan Schist in 1904 was simply an impossibility. Modern TBMs and tunneling techniques are amazing. Yes the stations have to be much deeper but modern construction technologies make it more than reasonable. That’s why shallow subways being built anywhere in the world are almost nonexistent. One only need to see the cost overruns mostly for mitigation on the Big Dig as well as the pictures I’ve seen of Manhattan when they were building the original subways in 1904. You simply cannot get away with that level of disruption today. Going under the services and just dealing with moving them at the stations is much more reasonable than along the entire route.

  • Newtonmarunner

    I’ll give you cut-and-cover is more disruptive. But not more expensive. Cut-and-cover also takes less time to complete a project than other construction methods. That’s why most places around the world do at least their stations cut-and-cover.

    I’d add that having been on the 86th St. SAS station three flights of escalators underground with a full-length mezzanine, SAS Phase I could have been done a lot cheaper.

    I’m also with Larry in that I like the shallowness of the 72nd and Broadway Station over the unnecessarily deep 86th and 2nd Ave. Station. It takes 10 seconds to get out of the Broadway Line stops in contrast to the 5 minutes it takes to get out of the new 2nd Ave. stops.

  • BA

    This article is full of inaccuracies. NJT uses the East River Tunnels for non revenue service. That consumes the “capacity” that is identified in the article. The ERT are at capacity.

  • AC Tesla

    You can’t make a blanket statement like that. In some places cut and cover is cheaper I grant you. But it depends on both geological conditions and the services that have to be moved. Moving sewer, steam, water, eletrical and communications can be exorbitantly expensive and it almost never can be done without major disruptions. 100 years ago, they didn’t pay mitigation to businesses disrupted. Today, we do. It also depends on the depths required to safely go under the services.

    Get used to deeper stations. They are the norm today. NYC’s Second Avenue Subway stations and the #7 extension are at about 80 feet and the Eastside Access Station will be 180 feet down. Relatively shallow compared to some stations in London or the really deep stations in Russia. Admiralteyskayais is 282 feet deep and the Arsenalna Station is a whopping 345 feet deep.

  • Alon Levy

    NJT uses the tunnels reverse-peak.

  • Don

    The SAS stations have little water sprinklers that can spray onto incoming trains in case they arrive on fire. These have to be maintained. There has never been a train arriving into a station on fire, but now they are spending a lot on that investment. There are many other examples of these “silo projects”… the post-Sandy repaired tunnel under the East River that carries the R train was not repaired correctly. Now older train models cannot fit through the tunnel. But in order to save face and open the project on-time, they ignored this problem and now spend more money on over-time and other costs rerouting older trains on longer routes to get to maintenance facilities. It also makes the system less flexible.

    Over time, they are just making small little cuts that will kill us in the long-term.
    They have trouble getting things done between groups in their own departments –– so how are they going to plan intermodal connections? Or regional connections, through-running, PATH connections, etc? It’s not about the technology or the funding… it’s about the politics.

    There’s no good reason why SAS or other capital projects are taking so long. The original city subway (IND) was built quickly, and while it was more expensive than the private subways (IRT, BMT), it was certainly more efficient than today’s construction. By the 1930s, no one would tolerate any more els, so everything was built underground. Now, people don’t tolerate cut-and-cover, so it’s built deep underground, if possible – meaning deep mining at stations, which is expensive… but the actual TBMs should be cheaper, since they have fewerworkers, they’re not cutting up all the pipes further towards the surface, etc. Who knows why it’s still so cost-prohibitive, and downright tragic. If government made it a priority, it would get built quickly… wasn’t that replacement bridge in the Twin Cities built within a year after the original collapsed?

  • Kylo

    East Side Access doesn’t connect the LIRR to the Metro-North, perhaps because of steep grades and different third rail technologies (over-running versus under-running), but it at least shouldn’t take decades to get built. Fulton Center is right in Lower Manhattan, but it’s only a few stories tall and they could at least sell the air rights – perhaps they didn’t build taller because the MTA doesn’t have real estate development expertise, or because they didn’t know if Lower Manhattan would bounce back after 9/11, or because they didn’t want to compete with the WTC, which has always had vacancy problems.

    I guess a lot of these problems come down to incompetence, ignorance, mismanagement, poor leadership, lack of incentives… a lack of vision? The Mayor not caring about transit or congestion pricing? The Governor trying to look good, not admiting problems? Maybe it’s the electoral college, campaign finance problems, gerrymandering, but the system seems broken…

  • Albatros

    The two are related because the operating budget pays for the service you get, while the capital budget dictates the quality of that service.
    Costs in the operating budget, including health care, insurance, and fuel, have increased in recent years, meaning it costs the agency more to provide the same service.
    Shortfalls in the capital budget put additional pressure on the operating budget.
    Delays mean people miss out on interviews, dates, school, appointments, jobs…


Long Island's bus and rail networks are run as two separate and unequal systems. Photo: Adam Moreira/Wikimedia Commons

On Long Island, Transit Operates as Two Separate and Unequal Systems

The Long Island Railroad is building some of the biggest infrastructure projects in the region -- even the world. The hugely expensive East Side Access tunnel and terminal at Grand Central and the construction of a third track for the LIRR Main Line will open up new possibilities for convenient, all-day transit that people can use for all types of trips. But not if Long Island continues to operate its rail and bus networks as a two-tiered transit system.