DOT Will Replace East Village ‘Mixing Zones’ With Semi-Protected Intersections

An offset crossing at Fourth Avenue and East 13th Street in Manhattan. Image: DOT
An offset crossing at Fourth Avenue and East 13th Street in Manhattan. Image: DOT

The First and Second Avenue protected bike lanes are about to get a whole lot safer in the East Village.

Now that both Manhattan roadways have been repaved, DOT says install new intersection designs that enhance safety for people on bikes and on foot — culminating two years of pressure from  cycling advocates after the death of Kelly Hurley.

They’re two of the city’s safest protected bike lanes, but two years ago, a box truck driver turned into and killed Hurley, 31, at the intersection of First Avenue and Ninth Street. In the aftermath of her death, advocates implored the agency to rethink its use of “mixing zones” — which force cyclists and drivers to negotiate the same space at the same time.

In a mixing zone, motor vehicle traffic turns across the path of protected bike lanes. Image: Reed Rubey
In a mixing zone, motor vehicle traffic turns across the path of protected bike lanes. Image: Reed Rubey

After Hurley’s death, Upper West Side architect Reed Rubey came up with an alternative design, which was subsequently endorsed by Manhattan Community Board 4. Rubey’s efforts partly inspired DOT’s chosen solution: the offset intersection, which it piloted at select locations in 2017 and 2018. In September, DOT’s “Cycling at the Crossroads” report showed that cyclists felt significantly safer at intersections with offset crossings [PDF].

Yet while most DOT surveyed feel safer cycling at offset crossings, the agency found that the design tricks some cyclists into thinking they must yield to the driver, in part because drivers were turning at high speeds. The report recommended modifications to slow driver speeds, and DOT continues to use mixing zones on lower traffic streets.

After a slow initial rollout, DOT is going big on offset crossings. In addition to First and Second avenues, it plans to include them in Brooklyn’s Fourth Avenue redesign, which is being implemented.

Bike New York Communications Director Jon Orcutt, a former city DOT official, welcomed the revamp of the Manhattan avenues.

“Since DOT published its report on bike lane intersections last fall, it’s been an open question whether we would see the offset design primarily along new bike lanes or also mixing zone retrofits,” he said. “We applaud this important step.”

  • Joe R.

    Can we just not prohibit turns on to minor side streets? The problem here is not just how you deal with turns, but how often they occur. Granted, mixing zones totally suck and what’s shown above is a better design. However, jogging out at every other intersection would frankly be something I would find beyond annoying. My guess is others will, also. End result is they’ll ride in the same line across the intersection as before/after it, which means they’ll be about half inside that zone where vehicles stop (in theory) before turning. That kind of defeats the purpose of the entire thing. Now if the turns only occurred at major intersections every 10 blocks or so, the desired path for cyclists might actually be mostly followed.

  • SSkate

    A “whole lot safer”? Reserving my judgment until NYC_DOT actually stripes lanes, etc.

    Skated up First Ave late Tuesday eve. Saw a cyclist ahead of me almost get left-hooked in the recently repaved section of the ave, although that could be blamed on the car driver turning left on a red.

  • Jacob

    NACTO just released design guidance for protected intersections:

    Curious to see how this stacks up.

  • Simon Phearson

    These “offsets” cannot and will not work unless they are physically designed to prevent drivers from taking the turns too quickly. Full stop. Mere paint won’t cut it. Drivers must have to slow down and reorient their cars in the direction of travel if they are to be effective.

    That is why cyclists “yield” to drivers on a lot of these the DOT has just painted on the ground. We can’t see the drivers, the drivers can’t see us, and they plow through the turns.

  • ItsEasyBeingGreen

    The example pic isn’t high enough visibility for the turning driver. The one at Columbus Ave/W 70th St has green right up to the crossing point, which is a start. These should also have intermittent green across the cross street itself.

  • jeremy

    Cute that DOT thinks this design might work

  • Jacob

    They use the flexposts now, but they surely must spend a ton of time replacing them. Why not use something tougher, like a curb?

  • Great news for all cyclists… except me. I don’t have any problems with mixing zones. I actually like them, they give me the option to merge behind a left-turning car. I feel more in control this way, versus putting myself at the mercy of drivers, hoping they are paying attention. I wince everytime I see a rider dart in front of a car that’s about to turn. I understand that the rider has right of way. But I never liked putting myself in that situation. To me it feels insane to do so.

    Most of the time it works out fine and nobody gets hurt but still feels like too much risk, and too much trust being given to random strangers behind the wheel of fast-moving multi-ton machines.

  • qrt145

    I like mixing zones too. Split phase signals may be safer but they also slow cyclists down way too much. I haven’t tried the offset design, but from looking at it I suspect it will involve playing chicken with drivers. I’d rather just merge.

  • The best mixing design is on First Avenue near the U.N. The cars that are turning left off of First Avenue are meant to cross over the bike lane far from the corner; and bicycles can then continue straight on First Avenue to the right of left-turning cars.

    I would provide a link to Google Maps; but doing that got a previous post of mine hung up in “Pending” limbo. So I will just post an image.

  • AMH

    It depends on how much traffic there is. Too much, and an impermeable wall of cars completely blocks the bike lane. Too little, and drivers blow through the turns at speed. Just the right amount and you can wiggle through to get around turning vehicles without slowing down.

    I hate the split phase signals too, since they stop you about every 2 blocks (you can usually get 4-6 blocks per light cycle otherwise).

    Not sure how the new design will work, but it at least looks more similar to best practices.

  • So essentially you are advocating painted lines explicitly directing riders to mix, and then pass to the right of the left-turn lane? I suppose this is a good solution. Even though I don’t like this particular intersection by the way, I do like the photo, at least as a general illustration of why it’s better for a rider to be to the right of a left-turn lane.

  • I hear you on the impermeable wall. Usually I am asshole enough to squeeze myself in (since cars can only really form a wall at low speeds) but yes there have been times where you can’t safely merge in. In those rare instances, you can always stop at the intersection and wait until it is safe to proceed. Inconvenient sure, but safety first motherfuckers. That’s the beauty of a bike. If ever a street situation gets too hairy you usually have the option of shapeshifting into a pedestrian.

  • Well, the bicyclists’ path is a straight line; it’s the cars that have to move into the left lane (across the bike lane) in order to turn left.

    I certainly recognise that having such a crossing at all is less than ideal. But, if left turns are allowed at a given corner, then the cars have to cross the bike lane at some point. In that case, I prefer them to cross the bike lane at a point when everyone is going in the same direction and the sightlines are clear, rather than during the turn when the drivers are more likely to fail to see a bicyclist.

    There are many locations where this configuration would represent an improvement.

  • Ian Turner

    Curbs are orders of magnitude more expensive to install. Besides requiring a lot more equipment (because of concrete work), they also necessitate re-evaluating and potentially re-designing draining systems, underground utilities access, etc.

    A better question is, why not bollards, but I fear the answer is that it might have negative impact on the safety of bad drivers (by making them responsible for their actions).

  • bushwicken

    Kelly Hurley’s tragic death had absolutely nothing to do with mixing zones. The article – like many others on Streetsblog – omits as to conflate for politic gain. An offset crossing is an improvement here. Kelly Hurley should still be alive. An offset crossing wouldn’t have stopped a reckless driver from turning across 4+ lanes of traffic and killing her. Why mislead when the goal you advocate for is worthy on its own merit?

    And whoever the heck Reed Rubey is, anything he came up with definitely had no influence on the adoption or design of offset crossings by the city. Is this just a convenient claim of causality to give a guy a shout out? I doodled a rocket ship once in third grade and then years later NASA launched the mars rovers, and the press never once gave me credit. Those jerks!

  • walks bikes drives

    Why not raised crosswalks or other form of speed bump before the bike lane to cause a slowing of drivers?

  • walks bikes drives

    If you prohibit turns onto side streets, how do you drive down a side street?

    The jog out there is barely anything, a slight curve. That should be no problem.

  • Joe R.

    Here would be the entire idea:

    1) Pedestrianize all minor side streets, which basically means bollard them off at one end so they can’t be used for through traffic. The minor side streets will remain through streets only to pedestrians and cyclists, not to motor vehicles. The major cross streets will be the only way to get crosstown by motor vehicle.

    2) Put the entrance to the side street on the side of the avenue which doesn’t have the bike lane. For example, the side streets east of 1st avenue would be entered AND exited from the east side of first avenue. The side streets west of first avenue would be entered and exited from second avenue, and so forth.

    3) Once you do all of the above, traffic signals need not apply to either pedestrians or cyclists on the side of the avenue with the bike lane since by definition there can be no motor traffic crossing their path. Cyclists can get a flashing yellow yield to pedestrians when pedestrians get a walk signal to cross the avenue.

    4) This creates ~10 block stretches where pedestrians can walk totally unimpeded by traffic signals, and cyclists just have to yield to pedestrians crossing the avenues occasionally.

    5) You can make this idea even better by having bike overpasses at the major cross streets, giving cyclists a virtually nonstop trip, other than yielding to pedestrians as mentioned in #4.

    6) Breaking up the grid like this also dramatically decreases the capacity for motor traffic in Manhattan, which in turn should discourage driving.

  • Ian Turner

    For the same reason: Any change to the road surface requires tons more work than changes to street hardware.

  • crazytrainmatt

    DOT has been installing rubber speed bumps in place of flexposts for left turn calming. They have a variety of designs, but all are more durable than posts and drivers seem afraid to run over the beefier ones. The main problem is that speed bumps inherently dangerous for cyclists (e.g. overtaking or swerving to avoid a pedestrian) and therefore need to be both clearly visible and outside the ROW.

  • crazytrainmatt

    Now would be the time to widen the bike lane to accommodate the huge increase in traffic since it went in. Even a few feet would make a huge difference given the mix of ebikes, hoverboards, scooters and whatever else one finds these days.

  • Don’t forget about the oh-so-sensible e-unicycle. Been seeing disturbing numbers of those out on the street lately.

  • 7) All bicycles should be designed to hover safely 20-30′ above all traffic.

    I guess the point I am trying to make is that your suggestion of permanently closing off minor streets to through traffic strikes me as only slightly less unrealistic than my sarcastic bullet point. Which is not to say I don’t like the idea though. But we are dealing with a community who will throw glass shards on the ground, they are that opposed to a simple painted bike lane. All-out civil war will break out if side streets are ever bollard’ed for the sake of bicyclists IMO.

  • Joe R.

    Well, my idea of viaducts is actually pretty much the same thing. Those might actually be built if NYC were willing to spend serious money on bicycle infrastructure. They would overcome all the usual objections to bike lanes, plus they would make pedestrians who claim to almost get hit by speeding bikes happy as well. Sadly though, if it can’t get done with paint and plastic delineators, it doesn’t get done. We can only hope that changes with the next few administrations.

  • Daphna

    I do not like that design at all. Drivers do not stay in the lanes they are supposed to. They often straddle the lanes and travel at high speed and it does not feel safe having a painted bike lane sandwiched between two motor vehicle travel lanes with drivers who are speeding and making no attempt to stay within their lane lines.

  • Daphna

    What people perceive to add safety to a street and what really adds safety often differ.

    Residents often clamor to have stop lights added to an intersection wrongly assuming that that will help with speeding and failing to yield, but signals are not a good tool for those goals and often make the situation worse, but people still cling to the idea that adding the signal will help and clamor for it. The DOT then gives in to the community pressure and installs a signal that they know will not help the traffic conditions that residents want help with.

    Mixing zones versus offset crossing also seems to be an instance where there is the perception that offset crossing is safer, and people are asking for it, but the DOT statistics show mixing zones are actually safer, but are giving in to pressure and rolling out the offset crossings against what the statistics on safety dictate.

    crazytrainmatt makes a crucial point: the bike lane needs to be widened. That is a change that is definitely needed. The DOT should stop counting the 1′ curbside gutter of a bike lane as if it is part of the lane when that space is not rideable and is always full of debris and water.

  • I don’t really disagree with your critiques; drivers do indeed do what you are describing.

    But I like my chances of being seen a lot better when I am right in front of the driver’s face going the same direction, than when the driver is in the midst of the turn.

    Short of banning turns in the direction that crosses a bike lane (a left turn here and on several other Manhattan avenues; a right turn in most other cases), there is no ideal solution.

  • Simon Phearson

    Yeah, I can’t believe that anyone with any real experience riding in this city – and especially midtown by the UN – would recommend this design.

    Ferdinand’s assessment, I’ll wager, is based less on his experience riding through this intersection than it is some theoretical speculation about how drivers and cyclists should interact.

  • Simon Phearson

    My concern is that they won’t use flexposts, or use them only sparingly. The flexposts on the 13th Street bike lane, for instance, are so widely spaced as to be functionally useless. There are a number of painted and “buffered” bike lanes throughout the city where they chose not to use them at tall. There’s an intersection along the Skillman bike lane where it should function like an offset/protected intersection, but it doesn’t, due to the total lack of any physical constraint on drivers’ movements.

  • Simon Phearson

    I haven’t seen many of these speed bumps, but the few that I’ve encountered have precisely that problem, for me. I’m always trying to avoid slower cyclists or pedestrians at precisely the point where they’re located, so I end up veering into traffic lanes.

  • crazytrainmatt

    The bumps I’m thinking of are placed in the buffer between the green-painted bike lane and the parking lane on the avenues (to control left turns onto the ave). I’ve seen a few delivery guys on ebikes go right over them at speed but I sure wouldn’t risk it!

    Moving them further away from the bike lane might let left-turning drivers squeeze through the gap, so I’m not sure what the right answer is. They definitely make a huge difference for pedestrians in the crosswalk.

  • crazytrainmatt

    I agree completely with your points regarding perception vs. actual safety. But DOT has a mixed record with a data-driven approach. They have been most successful in using it to modify concerns about vehicle speeds. But if we argue they need to follow the data regarding bike lane design details, then the advocacy community needs to make a big push for more relevant data collection and analysis.

    For example, DOT’s evaluation of the new partially protected 2nd ave lane between 59th and 43rd concluded that it is working great, when blockage is a bigger problem on the new section than elsewhere. Their analysis of the 26th/29th lanes cited vehicle speeds and increased bike traffic, but there were no statistics on bike lane blockage (extremely frequent) or bike speeds (quite slow given the narrow lanes). The 2018 intersection study studied only two examples of the protected intersection design. I’d have to guess that traffic and turn volume on 1st ave, especially as you enter midtown, is higher than any of the locations where they have been installed so far, and it is exactly at these intersections where the mixing zones become so uncomfortable.

    And while it’s the details that make or break the design, DOT doesn’t move fast enough to have an A/B test for every possible detail. For example, 1st ave at 55th has a “delayed turn” where the protected phase is over by the time the cyclist green wave get to the intersection; the fully split-phase ticket traps on 14th and 59th are even worse. The 2018 study didn’t specifically assess the highest risk situation: non-compliance at split-phase intersections; this is rare enough that most cyclists don’t look for it and the queue in the turn lane restrict visibility for the occasional car that makes an illegal turn at speed from the through lane. Split-phase intersections have the major disadvantage that it is impossible for an overtaking cyclist with a grreen to assess the intentions of cars queued to the right (and indeed the cyclist must be watching the left for crossing pedestrians meandering into the bike lane).

    I think one concrete step we can all take is to ramp up submitting 311 reports. The Reported app makes this a piece of cake. Even if NYPD ignores them (but note that TLC will fine taxi drivers), these can form the basis on which to petition DOT for design improvements.

  • puregutsxc

    NACTO just made me cry, the first time anyone in America has actually got it right.

  • puregutsxc

    Honestly we should take a page (again) out of the Netherlands design book. Should make the lane solid green through the intersections, and use clear yield markings (saw-tooths/block markings) in the car lane as it crosses.

  • I don’t know much about the DOT and how they operate but their everyday lack of integrity is easy to spot. Their own printed guidelines specify that cyclists should merge into the vehicle lane in mixing zones, and NOT pass on the left.

    But out on the street there is no signage whatsoever to indicate this, and the only signage that does exist, says that left-turning drivers must yield to bicyclists, which clearly implies that bicyclists should hold their straight line and force cars to stop and yield to them. DOT could not be any more contradictory on this issue.

  • ERoberts

    For a second I thought that was an entire defined bikeway. If only.


Evidence That Split-Phase Signals Are Safer Than Mixing Zones for Bike Lanes

When DOT presented plans for a protected bike lane on Sixth Avenue, one point of contention was the design of intersections. How many intersections will get split-phase signals, where cyclists and pedestrians crossing the street get a separate signal phase than turning drivers? And how many will get “mixing zones,” where pedestrians and cyclists negotiate the same space as turning […]