Less Service on the L Train? Wring More Efficiency Out of the Streets

The morning commute on the Williamsburg Bridge on November 1, 2012, when Sandy had knocked out the downtown subway network. Photo: Elizabeth Press

Gothamist dropped a bombshell earlier this week: To repair Sandy-inflicted damage to the L train tubes between Manhattan and Brooklyn, the MTA will have to suspend service through the tunnel for large chunks of time.

The repairs can get done fastest if the MTA halts service around the clock, but that would still last one to two years, according to Ben Kabak at Second Avenue Sagas. The agency can maintain some service by doing the repairs one tube at a time, but that would drag out the process to at least three years.

Either way, we’re talking about a significant hit to transit capacity that will affect hundreds of thousands of people — on peak days there are close to 300,000 trips through the tunnel.

Mayor de Blasio told reporters yesterday that this isn’t the city’s problem since the state runs the MTA. That argument makes sense in many cases, but not this one. Regardless of how the MTA conducts the repairs and adjusts service on other lines, the city will have to play a large role in planning for this shock to NYC’s transportation system.

After Sandy knocked out a huge portion of the city’s downtown subway network, the city and the MTA teamed up to repurpose streets and bridges for high-capacity bus service, including buses that took riders over the Williamsburg Bridge. Some sort of service like that will have to happen again while the L train tunnel is repaired.

Tops on the list of street improvements has to be a Williamsburg Bridge bus lane, as Ben mentions in his post.

Other steps will have to be taken to improve non-automotive travel between northern Brooklyn and Manhattan — and much of this can stay in place after the repairs are done:

Better approaches for the Williamsburg Bridge bike path. A center-median protected bikeway on Delancey Street would enable more bike travel over the Williamsburg Bridge, and so would a bikeable Meeker Avenue, with short connections on Marcy Avenue and Rodney Street. DOT hasn’t included bike lanes in its Meeker Avenue safety plan so far — the L train disruption should change the agency’s calculus.

Bike-and-ride parking. For most L train riders, walking to the J/M/Z or the 7 won’t feel like a great alternative, but biking to those subways is very doable. Massive increases in bike parking will extend the accessibility of subway lines for riders stranded by the L.

Citi Bike. Beefing up bike-share capacity in the affected neighborhoods will take some stress off the system.

That’s a start. How else would you like to see streets change so people who count on the L aren’t stranded?

  • Simon Phearson

    Great ideas, but I hope the city doesn’t do this without ample CB input, particularly from far-flung neighborhoods where drivers that use the Williamsburg bridge live. We shouldn’t do anything without their sign-off.

  • Mackle

    Increase ferry service. More citibikes at ferry stops.

  • mfs

    Sure the city has to do the bus lane, but the MTA has to agree to run buses on them!!

  • Joe R.

    Fast track digging a third tunnel. Once complete, start repairing one tube at a time. In theory that should result in no disruption of service. Open question is if we dug the tunnel as fast as humanly possible, could it be done in a few months? I tend to think so.

  • kevd


  • Alexander Vucelic

    Core L Train stations are 2-4 miles from Midtown easy cycling distance

  • mikecherepko


  • In addition to bus lanes, mandate a vehicle occupancy of 2+ people during rush hours on the Williamsburg Bridge.

  • bolwerk

    This is what happens when, year after year, you let neoliberal car fans and borrowing scolds sucker you into believing rail is too expensive. There is no replacement that can meaningfully reduce the pain here, except rail redundancy, and it’s too late for that now.

    Any changes you make to other modes, however good they are in their own rights, are futile as meaningful substitutes for the L Train.

  • bolwerk

    I think we need more time than that. This is the problem with having rubber tire fetishists in charge of City Hall and the governorship.

    Sandy should have spurred a spate of subway and other rail construction (e.g., bridge light rail). Never you mind that at least three crossings now need or have needed this kind of major capital reconstruction, we’re still far from being safe from a repeat of Sandy.

  • Joe R.

    I’m thinking more of what this country could do post WWII than the sad can’t do nation it’s become now. Post WWII we probably could have dug a tunnel in 6 months using a bunch of guys with shovels. Now it takes 6 years to do the same thing with TBMs. Go figure.

    We’re not even close to safe from another Sandy. It WILL happen. It’s just a matter of when. Besides the things you mentioned, NYC should have gotten started on a major seawall project. It’ll be needed by the end of the 21st century, if not sooner.

  • bolwerk

    Not sure about post-WWII, but prewar tunnels usually took about 4 years. That seems in line with international practice now for sort of similar projects, maybe a little slow.

  • Joe R.

    I would just be curious how fast we could do a tunnel these days if we absolutely had to. That said, I agree with your general sentiment that anything we do at this point is like closing the barn door after the horses ran out. The MTA is going to face this issue over and over again in the near future as more of the subway damaged by Sandy can no longer wait for repairs. It’s a pity we really didn’t do proactive stuff like start building redundant tunnels not long after we drained the floodwaters from Sandy.

  • Yeah, I’d like to see bus-only lanes. And make a frequent shuttle bus that is free and drops people in Manhattan. Most of the people are going to have to take a second bus or subway to get where they need to go, so that cost will be close to nothing.

    Like very similarly here in Jackson Heights. The Q33 ends at 74th Roosevelt Avenue transit center – if someone gets on the bus going that way and doesn’t have enough $$$ on their card the driver almost always waves them on, knowing they’re going to need to put money on their card to continue anyway to get on the subway. If you watch the people who get off the busses 99% head into the subway terminal.

  • Jules1

    In theory this would be awesome, not just as a relief for Sandy construction, but as a long term way of adding peak capacity to the L train long after the construction is done. Not sure about the reality of funding and feasibility of construction though. With enough money you could expedite things with a pre-cast tunnel and a 24/7 work schedule, but it’s unclear if that would be quick enough.

  • BBnet3000

    Thank you for mentioning the possibility of a center median bike path on Delancey (which already has the necessary left turn bans and distinct phasing to make it to Allen Street). Allen/Delancey could also get a 4-way green for bikes, the first in the US.

    It’s too bad that this and a comfortable connection to Allen from the Manhattan Bridge are probably never going to happen.

  • Jules1

    Bike and ride options to the J/M/Z makes sense, it would be hard to squeeze more people onto any of the Queens subways during rush hour.

    This would be a good time for Uber and Lyft to step up ride sharing along the L train corridor as well.

    The key for all this to succeed though is to make sure the MTA does an expedited state of good repair blitz along the J/M/Z _BEFORE_ the L train shuts down, because we’re going to need an old subway line to run maximum capacity for a while.

  • Jules1

    Something the MTA does have control over would be running express bus service from Northern Brooklyn to Midtown via a Bus/HOV 3+ lane in the Midtown Tunnel.

  • Geck

    Or as part of a more substantial future Utica Ave. line.

  • AnoNYC

    And the NYPD has to keep traffic moving at intersections…

  • AnoNYC

    An educational campaign, explaining alternative travel options of all types would go a long way.

  • Larry Littlefield

    Turn the Essex St Trolley Terminal into bike parking.

    People could bike over the bridge and take the J/M/Z/F.

  • Nick

    What about the ability of using your MetroCard on the East River Ferry.

  • Alexander Vucelic

    we won WWII in 3 years

  • Larry Littlefield

    JMZ signals were rebuilt in the 1980s. The Willie B was rebuilt after that. The service was realigned in Lower Manhattan quite recently. To remove tracks.

    I wonder if they could go 30 TPH? The max is now something like 18. The could connect the Canarsie Line south of Broadway Junction right into the Broadway line to run another 12 tph to Manhattan.

  • ahwr

    The tunnel is less of a traffic problem than the roads on either end. bus/hov lanes approaching and in the tunnel would be good even without new bus service from Brooklyn. ~50% of people using the tunnel to get into Manhattan from 8-9am are on a bus, ~43% 7-10am.

  • Gilberto
  • sbauman

    The major question is why the 14th St Tunnel is maxed out at 20 tph. I’ve heard 4 different excuses.
    1. They don’t have enough CBTC trainsets (my theory).
    2. They don’t have enough power to run more than 20 tph (the current excuse).
    3. One of the ventilation towers is broken so one train at a time can proceed through the 14th St Tunnel (a previous excuse).
    4. The FRA has issued new rules that place a limit of only 1 train in a tunnel (this is a new wrinkle for the LIRR’s ESA – the MTA ran simulations to see if they could operate their advertised 24 tph).

    Let’s assume excuse #2 is correct. They currently have enough power to operate 20 tph in two directions (both tubes). Therefore they should have enough power to operate 40 tph in a single tube.

    If excuse #3 is correct, they have 18 months to fix or replace the ventilation tower.

    I believe excuse #4 is correct for Class I railroads. The subway system isn’t subject to the FRA. NYCT has severed all connections to Class I railroads to ensure they are not subject to FRA regulations. They also operate more than a single train in the Steinway tunnels. So, I discount it.

    If the ventilation tower problem isn’t really an issue, then they could operate a single tube up to 40 tph with the power constraint. I believe such operation would permit a sustained rate of 15 tph. The only time they currently exceed this rate is between 8 and 9 am, when they run 20 tph. Therefore, it should be possible to provide 100% of existing service for 23 hours per day and 75% of the service for the remaining hour. This is a lot better doomsday scenarios being presented.

    Here’s how sustained 15 tph operation can be accomplished with a single tube. The intermediate station service level capacity is 40 tph. This limit is due to braking, acceleration and dwell time constraints. The Moscow Metro operates 40 tph; the NYC Board of Transportation operated the Third Ave El at 42 tph. Brooklyn Rapid Transit figured out a way to design stations to eliminate dwell time; they operated 66 tph on the Brooklyn Bridge .

    CBTC is supposed to be able to squeeze every available inch of track to increase service levels. Let’s put it to the test and see if it lives up to its hype.

    Let’s start with all the tracks from 8th Ave to Lorimer St being empty. Start sending trains towards 8th Ave every 90 seconds (40 tph).

    The switches are located between Lorimer St and Bedford Ave on the Brooklyn side and between 3rd Ave and Union Sq on the Manhattan side.

    The running time between Lorimer and Union Sq is 7.5 minutes. If 15 trains leave Lorimer at 1 1/2 minute intervals, all the trains will reach Union Sq. in 7.5 + 15 * 1.5 = 30 minutes.

    The operation can now be reversed with trains leaving Union Sq. headed to Lorimer every 90 seconds. That operation will also take 30 minutes and trains will be able to head back to Manhattan again. That’s 15 trains in each direction in a 1 hour period. The only difference is that those 15 trains are running in only a half hour.

    Passengers would get used to the new schedule and adjust their entry accordingly. It’s better than any of the alternatives that have been suggested.

    The trick is what to do with the trains after that reach Union Sq. The bottleneck is the capacity of the 8th Ave terminal to turn trains around. It’s only 24 tph. The conventional technique would be to alter the center track between the 6th and 8th Ave stations to act as a relay track. Alternate trains would terminate at 6th or 8th Aves. Both terminals operating simultaneously would be capable of turning trains at 40 tph.

    The trains would line up on the Brooklyn bound track. This track can hold 10 trains between Union Sq and 8th Ave. The remaining trains would stay on the 8th Ave bound track and wait for the terminals to clear. This is a common occurrence.

    The last couple of trains would not bother to change onto the Brooklyn bound track. They would change direction at Union Sq and 6th Ave and wrong rail to the switch at the 3rd Ave Station.

    It’s going to take a lot of coordination to pull this off. That coordination extends to all operating aspects. Passenger movements to/from the platform have to be coordinated as well as train movements. It should be amusing to watch them try it.

  • Joe R.

    In Japan I think they would be able to pull off this level of coordination. In the US, forget about it. MTA personnel and culture just isn’t up to operating trains which keep to within a few seconds of schedule. Heck, they don’t even consider a train late unless it arrives at the terminus more than 5 minutes behind schedule. And compounding this are the passengers. For what its worth, there was a time when NYC subway passengers were probably close to their Japanese counterparts as far as loading/unloading quickly and efficiently. Now you have too many people riding around with their heads buried in their gadgets, or otherwise acting like they’re on some sort of pleasure cruise. And then you have the MTA’s ridiculous policy regarding sick passengers where the train has to be delayed until help arrives, instead of just quickly getting the sick passenger off the train, simultaneously calling for help, and leaving. You have a great idea. It’ll just never work here.

  • Bernard Finucane

    The lack of tunnels under the Hudson is the really bizarre part. You should be able to take a subway from Meadowlands to Columbus Circle, and an express train from Newark Airport to the World Trade Center.

  • Bernard Finucane

    A lot depends on station construction. In today’s NYC it makes sense to build a big amenity-rich underground station.

  • bolwerk

    Even if any of that other stuff is true, the explanation I’ve read elsewhere is they can’t reverse trains quickly enough at Eighth Avenue or Canarsie. For Manhattan-bound trains, they can turn some at Myrtle (and do, in fact), but there is no like this for alternative for Brooklyn-bound trains. A possible fix is building tail tracks west of Eighth.

    FRA doesn’t regulate subways.

  • franco45

    Express bus service drown Wythe and Driggs avenues and along Metropolitan Avenue feeding to an express lane on the Williamsburg Bridge with dedicated Delancey street lanes. Two lines are needed-one going to Wall St area the other going up 6th avenue with stops at major cross bound streets from Houston north to 57th. Brooklyn bound buses go down 5th ave. another line heading north on Kent/Franklin and thru the Midtown tunnel. A less expansive version should go in place now to alleviate current overcrowding and allow for easy expansion when the L train goes out of service. Get people used to the buses quickly. Hey at least we’ll be able to access the Internet as the MTA has decided the the one train line carrying the largest amount of tech workers should be e last to get internet access below ground.

  • bolwerk

    Sure, if your goal is to maximize the cost of every project. We can get probably several stations for the price of one Hudson Yards station if we don’t do that.

  • bolwerk

    Not that redundancy eliminates misery either, but it at least meaningfully reduces it.

  • sbauman

    The fix for a terminal that cannot turn trains sufficiently fast is to have 2 terminals (usually on separate branches.) As you noted Canarsie bound trains can be turned at either Canarsie or Myrtle Ave (they used to be able to be turned at Atlantic Ave before the MTA made improvements.)

    There is a potential alternate terminal in Manhattan. It’s 6th Ave. There are 3 tracks between the 6th and 8th Ave stations. The middle track needs access from the 6th Ave station. This would allow trains bound for 6th Ave to relay onto the middle track, while allowing through trains to proceed to 8th Ave.

    Both terminals working in tandem would turn 40 tph.

  • neroden

    Agreed. That’s completely absurd.

  • neroden

    Well more accurately Russia won the European War in 4 years, while the US won the Pacific War in 4 years — but the point remains the same.

    By contrast, the Sino-Japanese aspect of World War II started in 1931 and didn’t finish until 1945, so it was a 14-year war (longer than Hitler’s war).

    We don’t even seem to be able to get anything built in 14 years.

  • MatthewEH

    Build a variant of one of the new bike-ped-only bridges that Sam Schwartz proposed back in 2012. He was saying build a bridge from Hunterspoint to Murray Hill, with a spur over to Greenpoint on the east side; let’s flip that around and build a bridge from Greenpoint to Kips Bay with a spur over to Hunterspoint.


    If we started building it now and put some juice behind the construction effort it could be done before the L shuts down.

  • Andrew

    There is already a subway route from Williamsburg to Midtown without the help of the L: namely, the G to the 7/E/M. A bus would be slower and far far more costly to provide. And there is capacity available on the G and 7 and M (less so the E).

    The big challenge is getting people to the 14th Street area, not to Midtown. Perhaps a bus is a part of the solution, but even a very frequent bus can only possibly be a small part of the solution, simply due to relative capacities of buses vs. trains.

  • Andrew

    In addition to needing much more time, the interlocking work alone at either end of this proposed tube to tie it into the existing line would be seriously disruptive to service. Not 2-3 years disruptive, but quite disruptive nonetheless.

    And once we’re toying digging new tubes, is this the best place for one or should we instead add one somewhere else to provide new connectivity on a different route, permanently?

  • bolwerk


    Would have been nice to see public debate about this in 2012. My personal preference is to find ways to balance servicing transit frontiers with offering some redundancy.

    I’ve seen the Worth Street proposal suggested, but that probably doesn’t have a good way to intersect the L.

  • Brian Marks

    Increased G and M service would be the primary things that the MTA should implement. 24/7 full length service on the M is an absolute necessity, and in addition to the G and M lines, I would also have two special SBS bus lines, with each terminating at Lorimer Street, one heading westbound to Manhattan, the other eastbound to Myrtle Avenue or have some runs of the B44 and B46 SBS head into Manhattan. Ferry service also would be needed.

  • ahwr

    If that new bridge were to double bike traffic from Brooklyn to Manhattan it wouldn’t replace a tenth of L train riders entering Manhattan 8-9am.

  • MatthewEH

    Heck, a full L train closure would double that figure of itself, even without a new bridge. Don’t neglect the extra pedestrians either.

  • bolwerk

    What is the point of Myrtle-Lorimer bus service? The L isn’t going away on that route.

  • Alexander Vucelic


  • Vooch

    Bike and Ride is critical.

    Citibike Stations expanded by Fall 2018 to
    Ocean Hill
    East New York

    Joe R and I rough calculate that the distances are such that 20,000 to 35,000 Bike to Ride trips might happen with adequate Support. The distances from the L train catchment basin to M & G alternative stations is typically less than a mile – only a 6-8 minute Citibike ride.

    How many buses need to run to serve 35,000 trips ?

    Bike and ride offers a compelling mobility solution at nil cosf.


Brewer to DOT: Start Looking Into a Bus-Only 14th Street

Manhattan Borough President Gale Brewer is calling on DOT to study making 14th Street a bus-only thoroughfare while L train service is disrupted during Sandy-related repairs. To allow for urgently-needed fixes to the L train tunnel, the MTA is considering either a full shutdown of service between Bedford Avenue and Eighth Avenue for 18 months, or a three-year […]

Keep L Train Passengers Moving With Great BRT

The news that Sandy-related repairs will require closing one or both directions of the L train under the East River (the “Canarsie Tube”) for one to three years has understandably caused panic among the estimated 230,000 daily passengers who rely on it. Businesses in Williamsburg that count on customers from Manhattan are also concerned about a […]

Maloney: Use L Train Shutdown to “Upgrade Our Bus Service”

New York City should use the impending L train shutdown to make long-term improvements to bus service, U.S. Representative Carolyn Maloney said at a public forum last night. To repair the Sandy-damaged L train tunnel under the East River, the MTA will either close the subway line west of Williamsburg entirely for 18 months or run it at 20 percent […]

De Blasio Sounds Prepared to Let the L Train Crisis Go to Waste

The impending L Train shutdown should be a blessing in disguise — the impact of losing service west of Bedford Avenue for 18 months is so great, there’s no good option that doesn’t involve carving out lots of street space for buses, biking, and safer walking. Major redesigns of 14th Street, the Williamsburg Bridge, and the streets connecting to […]