DOT Proposes Flushing Ave Bikeway in Prelude to Major Greenway Push

Flushing_bikeway.jpgImage: NYCDOT [PDF]

Here’s a look at the Flushing Avenue bike path concept that NYCDOT presented to the Brooklyn Community Board 2 transportation committee last night. This project would add another preliminary link to the path of the Brooklyn Waterfront Greenway, following in the footsteps of the Kent Avenue bike lane. After a round of questions with DOT’s project team, the committee passed a unanimous motion to endorse the concept.

DOT is aiming to implement the new bike path in July, and Brooklyn greenway project manager Ted Wright told CB 2 members to get ready for more greenway planning in the meantime. The agency is holding a series of public workshops, starting next week, for the full 14-mile length of the proposed greenway, part of a master planning process that officials expect to run through 2012. The first workshop, open to anyone who wants to come, will take place at Brooklyn Borough Hall on March 25. (You can RSVP with the Brooklyn Greenway Initiative, which is sponsoring the workshops with RPA.)

The Flushing Avenue project would construct a two-way bike path from Williamsburg Street West to Navy Street, separated from traffic by a nine-foot planted median. Vehicle traffic would travel in one westbound lane, between two lanes of parking. Only three curb cuts providing vehicle access to the Brooklyn Navy Yard would interrupt the bike path along the length of the project. Passengers on the B69 and B57 would disembark at bus bulbs constructed in the center median, with eastbound bus routes diverted to Park Avenue.

Despite the current tendency of motorists to speed on Flushing, the high volume of trucks, and the absence of a bike lane, more than 300 cyclists ride there on summer weekdays, according to DOT counts. "People are already using it for recreation and commuting purposes," said DOT Bicycle Program Coordinator Josh Benson. With the recent completion of the Sands Street bike path and the Kent Avenue path, the attraction of Flushing as a bike route to the Manhattan Bridge and Brooklyn Bridge Park is expected to grow substantially.

Committee member Mike Epstein said the plan provides "a top-notch protected bikeway" on a street in desperate need of safety improvements. "It’s important to look at this as a traffic-calming project" that will benefit pedestrians too, he said.

Epstein also requested better bicycle connections from Vanderbilt Avenue, where the bike lane terminates south of Fort Greene, and Bedford Avenue, where the erasure of a 14-block bike lane segment has left cyclists without a continuous route to the Williamsburg Bridge.

Benson said DOT would look into both ideas, but that adding a connection from Bedford would be hindered by ongoing construction work on Flushing Avenue. The segment of Flushing between Bedford and Williamsburg Street is still being reconstructed, he said, because the Department of Design and Contruction encountered an underground stream.

After the presentation and the unanimous vote in favor, I caught up with Milton Puryear, head of the Brooklyn Greenway Initiative. Following last year’s drawn out fight over the Kent Avenue bike lane, this early vote on the greenway precursor for Flushing Avenue was a friction-less affair. He was smiling almost the whole time we talked.

  • JK

    This two-way, curbside, separated lane would be ideal on Central Park West and Riverside Drive in Manhattan, where there are very few turning movements. Of course, you’d have to remove a lane of free parking on both those streets. So, we can put that’s an idea that can get in the countdown queue just behind the second coming, the full Second Ave subway and congestion pricing.

  • Moser

    Actually this design relies on the parking lanes to protect the bikeway and to keep the active roadway narrow. Its effect on vehicle speeds on Kent Ave has been salutory. Translating it to these Manhattan corridors would rely more on one-waying the streets, which itself would be a significant undertaking.

  • J

    Great news!

    This would work so many places. I like the Riverside idea, JK, and I don’t think you’d even need to remove parking (which would anger locals). Instead, you could remove a travel lane, which would have a traffic calming effect on a road with a lot of speeding. Everyone wins.

  • This is a great improvement.

    One question: Did the DOT explain the logic for keeping the parking lanes at 10′? 8′ is a perfectly acceptable width, and would give more space (4′) over to either the bikeway or turn lane/median (assuming widening the sidewalk is not in the works).

    If the new parking lanes are not intended to be designated by striping, it would also narrow the effective width of the movement lane, slowing speeds further and decreasing the width for crossing pedestrians.

  • Alex

    It seems unfortunate that a ‘redistribution’ of the ROW which gives THREE times more space to motorists than cyclists is hailed as a revolutionary new development.

    Why not remove the lane of parking next to the median (as the median will still buffer the cycle track), reduce the other parking lane to 8′ wide and give 22′ more to pedestrians and cyclists? Each sidewalk could be widened by 3′ and then each bike lane could be 8′ wide. At this width, cyclists could pass each other without having to merge into oncoming traffic.

  • Mike, I assume the parking lanes are 10′ wide because there are many parked trucks on this street. 8′ may be adequate for parked cars but I doubt it is for trucks.

  • DOT said last night that the parking lanes are wider than normal in order to keep the moving lane relatively narrow, while allowing space for motorists to maneuver around double-parked cars.

  • JK

    My CPW, Riverside two-way fantasy was a curb bike-way swapped for the parking lanes on the East side of CPW (park side) and W. side of Riverside (River side.)The buffer/barrier would be two foot wide bollard buffer using six inches taken from the remaining travel lanes. The streets could stay two-way. I’m not holding my breath. My plan would require a velo-rution capable of removing 1500 or so free parking spots from the Upper West Side. That’s a lot of pitchforks.

  • Geck

    This looks great. I am glad to read they are looking into connecting to Vanderbilt Avenue (and with a little luck, eventually with Bedford).

  • Thanks for the clarification, Ben. The movement lanes could still be kept narrow without such wide parking lines, in fact, more so. But I see now that the reasoning is for the double-parked car condition.

  • Michael Steiner

    JK, I also like your CPW proposal as it would give (a) a south-bound path and (b) would get rid of all the annoying cabs and delivery trucks which stop in the bike lane even though their delivery is heading to the west side. But as J mentioned i also would think that CPW doesn’t need four travel lanes. I usually travel there in the morning north-bound and at least north of 86 the road is never at capacity. Not sure what would happen if (finally) the CP loop is closed. But for that one probably have a middle lane which is south-bound-only during rush-hours and otherwise a shared right-turn lane?

  • Passengers on the B69 and B57 would disembark at bus bulbs constructed in the center median, with eastbound bus routes diverted to Park Avenue.

    Has the city not learned from sixty years of failure that one-way pairs cut bus ridership?

    If the 9′ island buffer is so necessary, then lose a parking lane instead of screwing up bus service for everyone.

  • Andrew


    Much of Riverside already has only one travel lane in each direction. Unless you’re proposing to make Riverside one-way or to carve a lane out of parkland or the sidewalk, one of the parking lanes would have to go.

    Alon Levy:

    What’s your source that one-way pairs cut bus ridership? When I’m on the Upper West Side (which has both one-way and two-way avenues with bus service), I much prefer to ride the M7/M11 on one-way Amsterdam/Columbus than the M104 on two-way Broadway or the M10 on two-way Central Park West, since the M7/M11 is much faster. The traffic signal timing on two-way avenues is such that, each time the bus pulls into a bus stop, it loses the green wave and has to wait for the light.

    Even if traffic signal timing isn’t an issue, I don’t see why one-way pairs would cut ridership unless the alternative is bidirectional bus service on each of the two streets. But if the choices are two-way service on one street or one-way service on each of two streets, I’m not sure why ridership would be any stronger in one scenario over the other, unless one of the streets (the candidate for bidirectional service) has many more traffic generators than the other. Am I missing something?

  • Andrew: Jacobs mentions it in The Death and Life. Two trends in the 1950s cut transit ridership wherever they were implemented: whenever a streetcar was converted to a bus, or whenever a two-way line was converted to a one-way pair, ridership fell.

    The explanation for the superiority of two-way is that it requires people to walk less to the bus stop. For BRT/LRT, there are also the facts that two-way operation concentrates traffic generators and that good implementations will have signal priority anyway. Green waves are for transportation that has many alternative corridors and can’t expect signal priority, i.e. cars.

    Flushing and Park already have two-way bus corridors, so the proposed change would only worsen bus service. Clearly the plan understands that Flushing’s important enough to get bike lanes in both directions; it just shunts half the bus traffic to Park, located under the BQE for maximal rider inconvenience.

  • From Steiner Studios at the Brooklyn Navy Yards on Flushing across Kent to Franklin past Broadway Stages across the Pulaski Bridge to Silver Cup Studios in Long Island City on to Kaufman (“The Big House”) in Astoria Queens is New York’s film industry long overdue for a seriously safe wonderfully velo “Sunset Strip”.

  • Moser

    I don’t know that I buy the “walk longer” on one-way bus pairs. If you move a bus stop, it may get closer to where some people are going or coming from, not necessarily further.

  • jackson

    Hey JK
    Re: A separated side-path on Riverside Drive. Riverside Drive is a very popular route for the spandex crowd coming from Lower Manhattan and Brooklyn and going to the GWB, side paths get clogged and move slow if they are median and/or parking protected. Conventional bike lanes between the parked cars and the moving lanes, perhaps with painted buffer where the width is available, would be preferable.

  • Moser, this isn’t moving the bus stop; it’s a service cut. It’s taking two two-way bus routes and turning one of them into a one-way pair.

    But while we’re at it, the main traffic generator in the area is the Navy Yard, not the BQE underpass.

  • Actually, Alon, there are some very large housing projects along Navy St that might welcome having more buses stop adjacent to them.

    It’s unclear who rides those buses more, local residents (of whom there are many, mostly along Park Ave) or Navy Yard employees. I’m guessing it’s fairly equal.

  • All for Flushing


    The B61 and the B57 are not the same route at all, so that is not a service cut. The B57 continues down Flushing to Queens The B61 runs down park for a short period and turns north.

    The majority of people working in the Navy Yard work on the western corner of the site, thus they would still have access to the B57 along Navy.

    I agree that people might be inconvenienced with a walk to Flushing, but this is not a “service cut.”

  • Those Park Avenue projects already get buses, in both directions. Why would it be such a big deal for them to get a little bit of extra service in just one direction?

    And no, the B57 and B61 are not the same route. The service cut isn’t about combining them – it’s about eliminating one direction of bus service on a corridor.

  • Andrew


    If the eastbound B57 and B69 are moved to Park Avenue, they’ll be less convenient for people coming from Flushing Avenue and more convenient for people coming from Park Avenue. The bus that currently runs along Park Avenue, the B62, isn’t of any use to people going to Middle Village or Prospect Park.

    What Jacobs was referring to was parallel bus routes. It used to be possible to ride a bus down 3rd Avenue in Manhattan; now people on 3rd have to walk to Lex or 2nd to catch a parallel bus there, which increases access time. And Lex and 2nd already had southbound bus service beforehand, so this doesn’t make things any more convenient for them. But in our case, the buses on Park and on Flushing go to completely different places; moving the Flushing buses to Park would increase access time for people on Flushing but decrease access time for people on Park.

    There’s also been a half-century of experience and of research since Jacobs. What Jacobs wrote in 1961 might seem plainly obvious to you, but it’s clearly not so obvious to several of the commenters here, nor is it obvious to the planners at numerous transit agencies. What Jacobs presents is a logical argument to support a then-recent trend, not a scientific study. That’s pretty weak. Bus ridership was dropping in the 50’s because people were moving away from the buses. Was the conversion of two-way streets to one-way a contributing factor? Perhaps, but it certainly wasn’t the underlying cause.

    Finally, you once again bring up the mistaken belief that buses will magically turn all traffic signals green. That’s simply not practical in New York (will pedestrians have enough time to finish crossing between the time the signal detects the bus and the bus arrives at the intersection?), nor is it how traffic signal priority is implemented on the Bx12 SBS, nor is it how traffic signal priority is implemented on any other bus system I’ve ridden. Traffic signal priority generally means that a green light will stay green a little bit longer if by doing so it will allow a bus to make it through the intersection, and/or a red light will turn green a little bit sooner if there’s a bus waiting to enter the intersection. One-way progressive signal timing is going to improve bus speeds whether or not traffic signal priority is in place. And the B57 and B69 aren’t proposed for SBS anyway!

  • Have you ridden bus systems in countries outside the US? If not, then your experience is too limited to be relevant. There are cities in the world that successfully have signal priority, many of which have large pedestrian volumes. There are also cities in the world with signal priority on all local buses, including the ability to turn a red light green when they arrive. If New York’s planner community is too incompetent to replicate this, then the city needs better planners, not one-way pairs.

    If there’s been a half-century of research showing one-way pairs work better, then I’d appreciate links. This is not what’s been observed in Manhattan. The issue isn’t that bus ridership dropped; it’s that there was a large immediate drop every time a two-way route got converted to a one-way pair. This happened independently of the trend toward lower ridership.

    One-way progressive signal timing works only as well as the bus can stick to a schedule. You can’t make the bus run within seconds of a schedule, which is what a green wave would require. Even Geneva and Zurich go for 1.5-2 minutes, much longer than a signal cycle.

    The problem with saying that Park access time is mitigation for Flushing is that good transportation planning doesn’t work this way. If there are enough people on Park to overwhelm the Flushing demand, then the appropriate thing to do is move all buses to Park. If there aren’t, then the appropriate thing to do is keep the Flushing buses on Flushing. Think of it as the analog to the following problem: you have 100 clients in city A, and 200 in city B, and need to build just one factory serving all 300.

  • Kaja

    Meanwhile, Alon provides no links, and addresses none of Andrew’s argument otherwise.

  • No links to what, Kaja – Jacobs’ book? Pardon me for referencing dead tree publications. In the future I will ignore every piece of knowledge that isn’t available on the Internet.

  • Andrew

    Thanks, Kaja.

    Alon, the burden of proof is on you. You made the claim. I agree that replacing two-way service on each street with one-way service on each street can have the effect of reducing bus ridership somewhat due to increased access times. But that’s not what’s happening here, and other factors may come into play to offset the increased access times, such as reduced running times.

    What do you claim has been observed in Manhattan, and can you support your claims? The busiest bus route in the NYCT system (and I believe in the U.S.) is the M15. The third-busiest in Manhattan is the M101. They seem to be doing OK despite the one-way streets.

    Progressive timing improves traffic flow. That helps buses as well as cars. A local bus can make several stops before hitting a red light; a limited can keep moving from one stop to the next. That generally doesn’t happen on two-way streets with closely spaced signals. (This has nothing to do with the details of the schedule.)

    I’ve ridden bus systems in three continents, and I’ve never seen the sort of traffic signal priority you describe. I don’t think the sort of traffic signal priority you describe is possible, except under very controlled conditions. Pedestrians have to be given time to finish crossing the street before the light changes, and the effects on systemwide traffic flow have to be taken into consideration.

    Nuance, my friend, nuance. We’re dealing with complex issues here. Not everything can be reduced to extremes. One-way bus service isn’t necessarily the end of the world. Traffic signal priority doesn’t mean that buses will never ever encounter a red light.

  • Well, said controlled conditions apparently occur in Monaco and on the two-way arterials of Singapore. They occur on the streetcars in Toronto, more or less; the only way a streetcar encounters a red light in Toronto is if it has a stop just before the intersection, and if it takes longer than expected at the stop. (I can dig up the link for Toronto if you’re interested).

    What’s been observed in Manhattan is that bus ridership immediately dropped when two-way routes were replaced by one-way pairs. I gave the citation already. Now I’m asking you to extend me the same courtesy and cite the decades of research you claim that show it’s not a problem.

    The speed effect does not seem to exist. You mention the M104 versus the M11, so I checked the schedule in the northbound 72nd-110th section, to eliminate the difficult intersections further south and the two-way section on Amsterdam further north. It turned out that the speeds were identical, about 1.7 blocks per minute: from 77th to 110th, the daytime travel time on Amsterdam is 19-20 minutes, and from 79th to 106th on Broadway it’s 15-17 minutes. (Both intervals are the longest contained in the 72-110 section on the online schedules).

    The M15 and M101 have high ridership, sure. But the M15 would have high ridership no matter what: in half the Upper East Side it’s the only route. To claim it by itself as evidence of anything is like to claim that the cold winter in Siberia disproves global warming. The M101 similarly provides West Harlem-Upper East Side service, and runs two-way for about half the route anyway.

  • Andrew

    I’m not familiar with Monaco or Singapore; when I was last in Toronto, about 10 years ago, I recall seeing streetcars at red lights. Am I not remembering correctly?

    Google reveals plenty of literature discussing traffic signal priority. Most of them don’t even discuss traffic signal preemption (the absolute form that you favor); those that do dismiss it up front in bus applications. (Traffic signal preemption is more commonly used for, e.g., emergency vehicles.) Traffic signal priority makes adjustments to the signal timing to accommodate buses – e.g., extending the green phase at the beginning and/or end. Some installations give greater priority to buses that are running late to help them recover.

    Based on your descriptions, readers unfamiliar with Jacobs probably assume that she undertook a major study of one-way bus conversions around the world and concluded that one-way conversions, independent of any other changes to the transportation network, severely cut into bus ridership. In fact, she didn’t undertake a study at all. She states her hypothesis, with a few apparent supporting examples, and that’s that. She doesn’t even attempt to prove her point rigorously. All you’ve given us is a 50-year-old hypothesis. Would you care to present us with some data (preferably more recent than 1961) to back it up? You made the claim; the burden of proof is (still) on you.

    I stand corrected on the M104 vs. M11. The timetables certainly go against my intuition, but my intuition has fooled me in the past, and barring actual running time observations (and I’m afraid I don’t have the time right now to ride back and forth gathering data), the timetables are the best we have.

    To call the M15 the “only route” in half the Upper East Side is a bit disingenuous; every north-south avenue on the Upper East Side aside from Park and East End has bus service; the M15 has the M101/102/103 to its west and (as far as the low 90’s) the M31 to its east. And I don’t think the M101’s busiest section is on 125th Street or the other two-way streets.

  • No, you’re remembering Toronto correctly. Toronto’s streetcar signal priority system is pretty good, but it has kinks, especially when stops are located right before intersections. The signal priority there works by sensing when the streetcar approaches an intersection, and turning the cross-traffic light red after a suitable warning period. When the streetcar is cruising it works well, but when it has a stop the system has to estimate the dwell ahead of time, which doesn’t always work smoothly.

    The M15 is the only route in the 1st/2nd area. At Manhattan bus speeds, walking all the way to Lex for service to Midtown takes too long. Walking to York works only if you want to go to 57th. I don’t know what the busiest section on the M101 is. But even then, of the twenty busiest NYCT bus routes, only three – the M15, the M101, and the B44 – have significant one-way pair stretches.

    Jacobs doesn’t really rigorously prove anything. Hell, much of Glaeser’s research is about taking what Jacobs said about urban economics and turning it into mainstream economics (for example, see Glaeser’s explanation of New York City’s pre-Erie Canal rise in terms of economic geography and the neo-liberal notion of investor confidence).

  • Andrew

    Traffic signal preemption is not be considered a good design for bus (or streetcar) systems. Toronto doesn’t use traffic signal preemption; it uses traffic signal priority. See page 67 of this PDF for a description of Toronto’s TSP system. “Extensions are provided in 1-second increments from a minimum of 3 seconds until either the transit vehicle clears the intersections or a predetermined maximum green extension duration is reached. The maximum green extension provided by transit priority is 30 seconds, depending on the location.”

    Most cities – and most parts of New York City! – don’t have parallel north-south bus service on nearly every avenue. The M15 is busy because it’s in a high-density area a reasonably long distance from the subway, and it’s a fairly long route.

    I agree that Jacobs doesn’t prove anything. Her book was an excellent eye-opener, and I agree with much, probably most, of what she says, but a one-page (two-paragraph) discussion of one-way streets and buses is hardly what transit planners should be treating as gospel. If the question hasn’t formally studied yet, I think it would be an interesting study. And if it has, I’d be interested in seeing the results. My own intuition is that the effects of one-way conversions are probably not nearly as severe as you’ve made them out to be, especially where the new eastbound path doesn’t currently have comparable bus service already.

  • Aaron Whitby

    We had a part of Flushing by the Navy Yards one-way for two years while they ‘fixed’ the road, it was a nightmare for those of us who live here. Apart from the fact that it reduces bus service, it also pushes traffic onto Park which one would think is polluted enough already but seemingly the D of T think they need more carbon in their air. Also as trucks are barred form Park all the trucks servicing the companies in the side streets between Park and Flushing will have to turn around in tiny side streets. The schools, the residents, the businesses all inconvenienced so that a few people can ride their bikes between Billyburg and Dumbo (on a sunny day). The stats they use for bike-use in their plan are from June, I guess November through March wouldn’t have looked good enough…
    If they have to make a bike path kepp Flushing two -way and make the Navy Yards who cause almost all the day time parking to make on-site parkign for their businesses, they have the room..

  • So far, the online lit search I’m doing for bus route design only shows results for BRT, where the studies seem to unanimously regard two-way median-running buses as the optimal design. This mirrors existing worldwide BRT practice. I can’t find anything for local buses, though.

    The eastbound route would not provide too much for people because it would be effectively one-way. People on Park would not plan new trips around the Park route because the return would be on Flushing; those who are willing to walk to Flushing are already doing so.

  • Aaron, I may be only one person – but as soon as that bike lane is in place, I will be riding my bike (safely) to work in manhattan everyday instead of taking a bus, or my scooter.

    And yes, that will be on any day that there isn’t torrential rain or snow.

    I’m sure that, like every other business on a sidestreet, the S. Williamsburg businesses will be just fine. Truck-centricity on Flushing has only made the street a bleak, blighted Limbo in Brooklyn.


Tonight: Turn Out to Support Safer Biking and Walking on Flushing Ave

What it’s like to bike on Flushing Avenue now. Photo: NYCDOT We’ve got a late and important addition to the Streetsblog calendar: At tonight’s meeting of the Brooklyn Community Board 2 transportation committee, DOT will present a revised proposal to improve safety for pedestrians and cyclists on Flushing Avenue near the Brooklyn Navy Yard. This […]

Sunset Park Greenway: Big Challenges, Bigger Potential

A map of potential greenway routes and east-west connections in Sunset Park. Image: UPROSE A full crowd of about 60 people turned out for NYCDOT’s Brooklyn Waterfront Greenway workshop in Sunset Park last night. The meeting was the second of four sessions the city is putting on with the Brooklyn Greenway Initiative and the Regional […]

Can a Greenway and Two-Way Traffic Both Fit on Flushing Ave?

The greenway segment on Flushing Avenue would connect Navy Street to Williamsburg Street West. Image: Google Maps The current concept for the Flushing Avenue segment of the Brooklyn Waterfront Greenway footprint calls for converting the street to one-way westbound traffic flow. Two-way vehicle traffic, say DOT planners, will create conflicts that endanger cyclists and pedestrians […]

Tonight: Important Meeting on Flushing Ave Ped-Bike Safety Project

The two-way bike path concept for Flushing Avenue. Image: NYCDOT Brooklyn Community Board 2, NYCDOT, and City Council members Steve Levin and Letitia James are putting on a public meeting tonight to get feedback on the two-way protected bike path with planted pedestrian medians proposed for Brooklyn’s Flushing Avenue. The project is part of the […]