Recommendations for Minneapolis’ 36th Street Bikeway

For those not in the know, Minneapolis is proposing to construct a protected cycle track along  36th Street West this summer.  This project hits near and dear to me as my wife and I will be moving into a home a mere 200 feet from this facility in exactly two weeks.  Glancing over the design, this project seems like a total slam dunk.  I’m not going to spend my time here debating the merits of reallocating space, arguing if cyclists pay their fair share, or anything else. Minneapolis and St Paul (any city, really) need to roll out protected cycle tracks on key corridors whenever they get the chance, and this proposal links together the Bryant bicycle boulevard with the wonderful lake trails, while giving would-be bus riders some respect on what is among the worst pedestrian facilities in the city (also removing a pointless stoplight and consolidating some redundant bus stops – huzzah!).

Seriously. (source: Google Streetview)

But seriously. (source: Google Streetview)

With that said, it’s my personal opinion that we can do even better than the incredibly high return project as proposed today.  Conversations with project staff and city leaders (who have done a wonderful job thus far, and as always my critique is not an indictment of their work, passion, or political challenges they face) have keyed me in to a few assumptions that drove the current design.  I’ll take them on one at a time.

Parking Location & Goals

Between Dupont Ave and Richfield Rd, current on-street parking is limited to the north side of 36th St.  The bikeway plan intends to continue this design, using paint and plastic posts as the only protection for the cycle track while leaving on-street parking on the north side.  The rationale is that most properties and businesses along the corridor (Bryant to Richfield Rd) are on the north side, so keeping parking here is more convenient for street users.

In my mind, the convenience of motorists parking shouldn’t trump other potential benefits from alternate designs.  We know that barrier-protected bike facilities are safer than ones that use paint and plastic (even if a buffer with plastic posts is a MASSIVE improvement over the current design).  We also know that the northern sidewalk has built-in protection along most of the corridor to keep pedestrians feeling comfortable alongside moving vehicles:

Though some sections are a bit sparse on trees.

Though some sections are a bit sparse on trees.

We’re not talking about extra costs here, just moving the parking to the southern side of the street instead.  Bicyclists get the advantage of roughly 9′ of parked cars separating them from the moving ones, and southbound cars leaving alleys have a little better visibility to westbound traffic.  For those who like parking supply, this proposal also adds a few spaces since Lakewood Cemetery has no alleys and requires few turn lanes along the southern edge of 36th St.


12′ WB lane includes the gutter pan, which is admittedly more efficiently used under parked cars.

Bus Pullouts vs Bulbs

Project leaders have decided that the (fewer than current) bus stops along 36th Street will be pullouts, similar to designs across our region.  However, this pinches down the expanded pedestrian area and cycle track to shared space.  The net effect is basically asking cyclists to ride on the sidewalk.



While I’m all for implementing shared space zones (otherwise called woonerfs), that design doesn’t seem appropriate in this situation.  The cycle track is specifically designed to allow speedy bike travel, and mixing the two modes here seems like an unnecessary safety issue given the context.

Additionally, there’s something I’ve been mulling in my head recently regarding how we treat buses by forcing them to pull out of the traffic stream to keep auto traffic moving.  Traffic engineers have no problem imposing a stop sign or stop light on moving vehicles to let other vehicles enter the stream, causing some delay for thru-drivers (let’s say 15-45 seconds depending on light timing) perhaps 15 times an hour (depending on cross-street traffic).  For example, just a few blocks north along Hennepin, a light at 34th St halts southbound PM peak-hour traffic (683 vehicles per hour) to allow 18 vehicles per hour to enter the SB stream, 30 to cross Hennepin, and a few more to turn northbound (among other goals the stoplight has).  Why should we view waiting behind a stopped bus to be any different?  Especially when a bus could easily wait 4-8 seconds to re-enter traffic on a lane carrying 400-800 peak hour vehicles (longer if motorists are especially non-yieldy).

Like a temporary stop sign for vehicles

Like a temporary stop sign for vehicles

36th Street carries several buses, notably the 6 (west of Hennepin) and the 23 (east of Hennepin).  Let’s use the area on the southern side of the street designated for parking as bus waiting areas where necessary.  This keeps the perceived street width down while allowing buses to quickly stop, pick up passengers, and get rolling again.  Longer-term (read: not Summer 2014), when the street is re-constructed, the curb can be moved out to the edge of the cycle track and extended at bus stops, with full shelters, heating, and benches):

Short-term design is just paint on pavement, no shelter.

Short-term design is just paint on pavement, no shelter.

Turn Lanes

The design calls for keeping an left turn lane at Hennepin Ave, and an eastbound right turn lane at Dupont (though the plan recommends the latter as a shared bus pull-out + turn lane).  By removing parking for half a block west of Hennepin and “chicane-ing” the thru-lane, we can maintain this critical left turn lane.  For Dupont, I would recommend moving the bus stop to the far side of the intersection and adding a pedestrian refuge island between a narrow turn lane and the cycle track:

Looking east.

Looking east.  Again, use paint in the short-term.


Traffic flow.  That’s pretty much it.  Losing parking on the north side eliminates the ability to create the pseudo westbound right-turn lane at Emerson project staff call for, and as noted the bus bulbs will slow down traffic once every 10 minutes when a bus comes by (during peak hours).  Beyond traffic, extending the cycle track to Bryant means the loss of on-street parking from one street direction – approximately 25 spaces (if my Google Map extrapolating is correct).  This is potentially harmful since the area between Colfax and Aldrich along 36th St has some great local businesses that rely (partially) on street parking.  Of course, there are solutions to the issue of crowded street parking, notably extending meters into the neighboring streets to ensure available supply for patrons (plus, bicyclists shop and eat, too).

In any case, this project really is a major win for Minneapolis, its residents, and businesses, even as proposed.  If you think these recommendations make for an even better, livable, safe street that prioritizes pedestrians and cyclists, pass this post along to Lisa Bender and the project staff.  Thanks, everyone!

19 thoughts on “Recommendations for Minneapolis’ 36th Street Bikeway

  1. Matt SteeleMatt Steele

    This street could also use some refuge islands to facilitate bike/ped transversability. Much more efficient than using stoplights to facilitate crossing.

    We need more bus bulbs and fewer bus pullouts. Pullouts cause lateral jarring and also require harsher acceleration/deceleration to get in/out of traffic faster.

    Finally, space-efficient access and mobility (good biking and walking infrastructure) is a primary function of an urban street. This should be the default choice, and car storage can fill in any leftover right of way.

    1. Janne

      Matt, could you translate your comment into English for those civilians wanting to understand the implications of this proposal? Sounds interesting, but I have no idea what you said.

      1. Alex CecchiniAlex Cecchini Post author


        – Islands in the middle of the road similar to the ones cropping up on bicycle boulevards as they cross major streets would calm traffic and make crossing easier. For now they would have to be paint not raised concrete.
        – In addition to the logistical/equity issues I cited, bus pull-outs are bad because they make buses feel herky jerky. Like playing jello in the backseat of a car.
        – Bus lanes, bike lanes, wider sidewalks should take priority over on-street parking in most cases.

          1. Alex CecchiniAlex Cecchini Post author

            Sure 🙂 Lateral = side to side. Bus pull-outs often require fairly sharp turns, both entering and exiting. This makes for an uncomfortable ride with higher side to side g forces. As Matt also states, buses that need to re-enter traffic flow from a pull-out may jam on the gas in an effort to get into a tight opening. This means harsh front-back g forces.

            To the lay-wo/men out there, this is why bus rides are almost always less comfortable than rail (subways, LRT, streetcars). It makes people less likely to stand in a bus, which means more people will perceive a bus “full” when all seats are occupied (whereas people will gladly strap-hang on a subway).

            We need strategic investments in rail, but our bus system does and will continue to do the heavy lifting for the majority of transit riders out there. Little things like bus bulbs that speed up service and make rides more comfortable make transit (in general) all the more attractive and should not be ignored or downplayed when talking street design.

  2. Sean Hayford OlearySean Hayford Oleary

    One minor problem with your proposal is that W 36th St is an MSA route, and subject to state standards. Unfortunately, two- and three-lane MSA routes are required to meet state aid standards for lane width and “curb reaction distance”. When you have a travel lane directly next to a curb on a 2-lane road, you need 4′ of curb reaction distance. Whereas if you have a parking lane, you don’t need any of that extra space. So, if you switch parking to the non-curb side, you lose that extra 4′ to work with.

    Personally I think the original proposal makes more sense. Parking-shielded bikeway may be marginally more pleasant, but I don’t think a typical user will find the bollard-defined buffer too abrasive. All things being equal, it does make more sense to me to have parked cars on the side of their destination.

    I do definitely agree with you on bus stops. Pullouts are a waste of space — and worse, limit the speed and efficiency of bus service. The buses on 36th and Richfield Rd only run every 15-30 minutes, and there are no time points on 36th itself (where the bus might have to wait if it’s ahead of schedule), so the maximum likely delay is like 30 seconds.

    1. Janne

      Thanks for this post! I’ve been to comment on it, but hammy Gf time to understand the details. I’m notified about the bus stops. One of the worst place on the Loring Bikeway is where the bus loading zone is in the bike lane. Riders are unaware bike are, coming and it’s dangerous for everyone.

    2. Alex CecchiniAlex Cecchini Post author

      Thanks for the comments, Sean. Couple thoughts/remarks:

      Not arguing with you, but the stupid rules. 1) where was this MSA requirement when one of 2 eastbound lanes ran right along a curb that contained a 3′ wide sidewalk that facilitates bus stops? This is just an example of a top-down system of rules that creates wider and more dangerous streets out of context, and needs to change (or be removed entirely).

      We’ll have to disagree on the parking protected vs plastic bollards. For most casual cyclists (including myself), the physical protection (coupled with good intersection design) is definitely an upgrade. At a minimum, you have 9 extra feet of empty space between peds/cyclists and moving traffic. Plus, bus bulbs on the southern side can only exist if parking switches there, while the north side of 36th has an ample boulevard space to place bus waiting areas. So while I agree that all things considered equal you’d want parking on the property side, to me the tradeoff is definitely worth it.

      1. Sean Hayford OlearySean Hayford Oleary

        When there are two lanes going the same direction, the distance required is less (2′). I suspect the current curb-side lane is either 11′ or 12′ + 2′. (Width is measured from the curb face, not the gutter seam.)

        The widened curb on the south side is not really a sidewalk, of course. This is basically in-line with a historical era where sidewalks were an amenity for a home or business, not a part of street infrastructure. A cemetery does not require that amenity, so it has no sidewalks. In theory I like the thinking (sidewalks as amenity) but in practice, it’s pretty crappy, especially when bus stops are in the mix.

        However, even if it were a sidewalk, I think it is telling that MSA standards are filled with rules about numbers of lanes, widths of lanes, required reaction distances, and yet not a thing about sidewalks or bikeways. It is 100% permitted to build a 4-lane urban street with no sidewalks at all on the MSA system, with no variances. But building a street with 10′ lanes, plus sidewalks and bike lanes — that will have to go through a difficult variance process.

        I don’t see why the bus bulbs are necessary if there is no parking on that side. Don’t bump in or out for the bus stops. Just have the bus stop in the lane, next to the posts. Pedestrians will have to cross the bike traffic, but I don’t think that’s a fatal flaw — even in Copenhagen, bicyclists have to stop and yield to transit users at most but stops.

    3. Walker AngellWalker Angell

      I think average users and others will find a great difference between cars or bollards as separators. Cars provide a much greater real and perceived element of safety. Cars provide a bit of noise and blow (air pressure from moving vehicles) protection. There will also be less road debris on the path with car separation.

      1. Cameron ConwayCameron Conway

        I’d agree, although a parking-protected bikeway does run the risk of passenger-side doors opening. Unlike folks coming out of the driver’s seat, passengers are conditioned to be opening their door toward a sidewalk… not towards a bike travel lane of ~15mph.

        1. Walker AngellWalker Angell

          There are a few things in bicyclists favor. First is that there are many fewer passengers than drivers.

          If the track is one-way then the bicyclists are generally riding on the right of the track and therefore farther away from opening doors.

          If the track is two-way, the bicyclists closest to the doors are usually coming from the front of the car so hopefully the passenger will see them coming before they open the door and if they do open the door the bicyclist will hit the front of the door which is much better than the back both in the shape of what they’re hitting and in the likely angle of the door and there is a possibility that the bicyclist will push the door closed somewhat which will also lessen the impact.

        2. Alex CecchiniAlex Cecchini Post author

          In addition to Walker’s thoughtful comments, the parking protected 2-way cycle track design above leaves the 3′ buffer zone the city proposed between traffic and bikes. 3′ may not be ideal, but leaves very few circumstances where the door would extend far into the cycle track.

  3. Suzanne Rhees

    Honestly I’m a little concerned about the protected cycletracks with plastic bollards as a separator. It seems to me that it makes cyclists less visible to turning vehicles. Noticed this the other day on Portland Ave. downtown, where this type of separator is currently in place (perhaps as a test?)

    1. Alex CecchiniAlex Cecchini Post author

      Hi Suzanne, and thanks for the comment. Could you expand a bit on the visibility piece? It seems that visibility for/of cyclists seems to be a major selling point for the thin plastic posts as they’re fairly infrequent and about half the height of an upright cyclist. Just curious for more detail.

      1. Suzanne Rhees

        Sure — I’m always going straight in the bike lane, where many vehicles are turning right onto E. Grant Street and again onto 15th St. to cross the freeway. It seems to me that I’m less visible to the turning traffic than if I were out in an unshielded bike lane. Similar to riding on the sidewalk and needing to stop at intersections. Again, it’s just a perception at this point and I’m interested in seeing how the cycletracks take shape.

  4. Cameron ConwayCameron Conway

    I love how you used a photo from Dexter Ave, that was my old commute in Seattle! I also think it’s an incredibly useful comparison in context. The bus stops on Dexter do cause traffic to grind to a halt and this is a really great step towards road-share equality. Considering that Dexter has two high frequency bus lines that each run with ~15 min headways, 36th looks like total overkill in comparison.

    Plus, on a street that has such little bus traffic? The current design is effectively giving priority to those few people who might have to wait 30 seconds tops, while forcing cyclists and transit riders to deal with multiple detours at all times of the day. Thank goodness that the concrete isn’t being poured in now, I want a far superior design before we make this permanent.

  5. Nathan Campeau

    As a resident of the 3500 block of one of the streets the cycle track will cross (welcome to the neighborhood, Alex!), I think that the current proposed design is superior to one that has a hard edge like a curb or parking. I work in the West Metro and come home on 36th eastbound. I need to be able to leave the cycle track and take a difficult left turn at rush hour (on a hill, no less). I have yet to see a design that will improve that connection for me (I’ll only be on the cycle track for 3 blocks), but it’s probably not worse than the way things are right now. I’m excited that this will make accessing the lake easier for neighborhoods to the east, but I do wish the design had more to offer people who will only be using short segments of the cycle track.

    1. Alex CecchiniAlex Cecchini Post author

      Great thoughts. I agree that a barrier protected (concrete or parking) cycle track definitely has the downside for existing at any time. In fact, except for at major intersections with signals (only Hennepin, Dupont, and Bryant moving forward), cyclists behind concrete are effectively treated like pedestrians if trying to access the smaller residential streets (need to stop, then wait for a gap in traffic to cross rather than being able to make a moving left from a general lane). This is definitely a tradeoff I didn’t discuss and it rarely seems to come up when cycle tracks are proposed.

      On the flip side, I don’t know how comfortable most general riders (the “8-80 crowd” if you will) would feel crossing through the paint and plastic into a general lane to make a left onto a side street rather than just waiting it out. I’m not sure on the solution, to be honest, other than to say that a better street design would hopefully have calm enough traffic that crossing is a) short in distance, b) easy to do as motorists can see/slow down, and c) comfortable enough that short stints like the one you describe could be done by a majority of cyclists if they choose to do so without “interrupting” other vehicle traffic. (note: I say that last part slightly in jest, but it is fair to be aware that this street will already see volume reduction and dedicated infrastructure for cyclists). Anyway, sorry for the lengthy reply, and great points.

Comments are closed.