SDOT’s traffic engineers are trying a different bike lane design at 42nd and Roosevelt after several people were injured at the U District intersection.
The intersection design is especially tricky because of the fairly long and steep downhill that allows both people biking and driving to pick up a bit of speed before the intersection. People driving are supposed to yield to anyone in the bike lane before turning, but far too many fail to do so.
“It has been brought to our attention that interactions between people driving, walking and riding bikes at the intersection of Roosevelt Way NE and NE 42nd Street are not working as intended,” SDOT wrote to neighbors recently via a door hanger note. “As a result, we are changing the way right turns are made.”
The newest attempt to fix the issue uses a design in the national NACTO guide on bike lane design, which includes a “combined” turn and bike lane. People turning right are now supposed to merge with the bike lane before turning. Workers installed the new design over the weekend.
Initial observations suggest many people are still turning across the bike lane incorrectly, though a period of adjustment is often necessary for road design changes.
Spent a few mins watching cars ignore new turn lane @ 42nd & Roosevelt. 4 of 5 cars turned from travel lane! pic.twitter.com/H494qIyMci
— NE Seattle Greenways (@NEGreenways) April 25, 2015
The city installed this short section of protected bike lane between NE 45th Street and the north approach to the University Bridge in January to demonstrate the planned bike lane design that will be part of a large repaving project scheduled to start later this year.
Whether collisions are reduced or not (we desperately hope they are reduced before anyone else is seriously hurt or worse), the “combined turn lane” design is never going to be super comfortable for everyone. Intersections and driveways are the most dangerous places in a bike route, and where people need protection the most. The combined lane is a low-cost solution that also removes much of the value a protected bike lane creates: The sense of safety and comfort for people of all ages and abilities.
A more complete solution would involve calming traffic speeds and creating a “protected intersection” design commonly seen in Europe but not yet tried in Seattle. This would essentially involve placing a new curb further into the intersection that would slow turning movements, create better sightlines and provide more space for people to avoid potential conflicts. The slight shift in the bike lane may also slow people on bikes, a good thing on such a long downhill. This video by Nick Falbo of Alta Planning + Design in Portland provides a great illustration:
Protected Intersections For Bicyclists from Nick Falbo on Vimeo.
Andres Salomon of NE Seattle Greenways has been pushing a similar design for Roosevelt and made this quick sketch of what it could possibly look like:
Note that was the less aggressive option to #FixRoosevelt. Other options for NE42nd & Roosie. @dongho_chang pic.twitter.com/usvkEGVtep
— NE Seattle Greenways (@NEGreenways) April 17, 2015
The advantage of putting a low-cost pilot design down on the ground in paint-only form is that the city can observe how their concepts work in real-time and make changes before installing a permanent design. That’s one reason they installed a short section of the Roosevelt bike lane early. And there is surprisingly little guidance on how to build safe bike lanes down big hills.
But they need to figure out how make intersections like this one work safely for all users. As of March, plans for a permanent design to be installed as part of the major repaving project later this year do not include this protected intersection concept (it’s a technical design document, so elements may be difficult to see):
Protected intersections are not the only options out there for making intersections safer. One option we have already seen would be to have separate signal phases for people turning and people biking like on 2nd Ave. Or, especially in places with very high numbers of people walking, the city could consider something like a “raised intersection,” which is like a big speed hump. This would greatly improve the walking and accessibility environment and slow people biking and driving.
On a street like Roosevelt, the biggest key to safety lies in slowing people down. A long one-way downhill with few major intersections, the street can feel more like a highway than a street in a very dense neighborhood. Anything the city can do to interrupt speeders would improve safety.
But what Seattle can’t do is repeat designs we’ve already tried that don’t meet all our traffic safety and bicycle access goals. Combined turn lanes simply don’t meet the “all ages and abilities” standard the city is supposed to be striving for in bike routes like Roosevelt. We know this because we have a ton of them all over the city, and people consistently report that they do not feel safe using them.
Safe intersections and driveways are the last big missing piece in Seattle’s bike lane designs. It’s not an easy problem to fix (certainly more complicated than I have laid out here), but that’s why we have some of the best traffic engineers in the nation. We have to figure this out.
Comments
30 responses to “City tries new bike lane design for collision-heavy Roosevelt intersection”
How about a speed limit for cyclists in the PBL (personally i wouldn’t feel comfortable faster than 10mph in that PBL), and an advisory sign for faster cyclists may use full lane?
Good ideas — the Dutch are doing that, according to CROW — too many complaints about fast bicycles making cycletracks uncomfortable for slower and more vulnerable users (children, elderly, cargo bikes, etc.) so the Cyclists Union has pushed to get faster bicyclists out of the cycletrack and onto the street.
It’s safer for the more vulnerable cycletrack users, safer and more convenient for the faster cyclists, and fine for motorists since the faster cyclists aren’t slowing down traffic. Initial reports say faster Dutch cyclists prefer the street in urban areas with high intersection density, but stay on the cycletracks in less built-up areas with low intersection density.
As a practical matter, though, Seattle Municipal Code has no provision for speed limits on bike paths, if the bike lane is considered an off-street facility, or for separate speed limits for bicycles if the lane is considered an on-street facility. (SDOT has signs in place on Broadway that say it’s a separate path, but I don’t know what the legal status of Roosevelt would be. Cycletracks aren’t really recognized at all by SMC or RCW, definitely a place where legal infrastructure doesn’t support physical infrastructure.)
I also agree this is a good idea. Two more points:
1. It makes it clear to cyclists that they don’t have to use the bike lane (I know the law doesn’t require it, but many people aren’t sure).
2. It might help make drivers more aware that it’s ok for cyclists to be in the vehicle lanes.
We should probably have all cyclists dismount and walk their bikes while in the protected bike lanes…I mean, really, why pretend bicycling is a practical alternative to the car?
Sarcasm aside, it’s true that 10mph is about the same as jogging. I’m fine with slow bike lanes for those who need them.
But there needs to be a route for fast riders. On some streets, that would be the regular traffic lane. In other places, such as Westlake N (along west lake union), I’m concerned that the planned bikeway will be inadequate. It is not safe to ride on Westlake with traffic going 35-50 mph.
This is a downhill route – pretty hard to go just 10 mph. However 10 mph is a 6-minute mile – a bit faster than jogging.
… and still wondering when any improvements will occur at the south end of the University Bridge.
Why these smaller intersections get attention, while the many bicyclists I see daily trying to merge across 3 lanes of speeding southbound traffic heading Harvard Ave E, up to Capitol Hill, is beyond me.
South of the University Bridge, the existing substandard bike lane is too narrow to even fit the full width of the bike lane marking, 40″ wide — much too narrow to meet the bare minimum width for a bike lane. So any improvement to Eastlake/Harvard would probably have to be a much larger project than just fixing the intersection itself.
Until then, I suspect the “official” position is that people who are comfortable riding in traffic can merge left into the left-turn-only lane, while others can ride past the intersection and the bus stop to use the crosswalk with the traffic signal.
What do you think of this short-term fix?
https://www.google.com/maps/d/viewer?mid=z0xXa0ZjM3qM.kgfrYsakr08E
Basically, with a new bike box on southsound Eastlake and a new left turn box at eastbound Fuhrman, people biking could more easily turn onto eastbound Fuhrman. There, new bike lanes could lead to a new all-way stop at Franklin, which could easily get the neighborhood greenway treatment and already has a traffic signal at Harvard (it does serve a trolly bus route, which would be rare for a greenway).
This would provide an easy and fast way to skip that horrible merge onto Harvard. Plus, the new bike box could help people get into the left lane ahead of that merge if they still prefer to go left on Harvard.
I actually prefer to make this maneuver today if there’s heavy traffic. But it’s not obvious if you don’t know the local street network well. Signage (<-- Broadway), bike lanes and better signal timing would all help make this more obvious. Obviously, a complete redesign of Eastlake is the preferred solution. But if we don't want to wait for that major project, this is a quick and cheap fix.
@Tom,
I like that a lot, and it mimics what I do already. I frequently take the University Bridge to Fuhrman/Boyer to bypass the U-District when I want to get to the Lake Washington Loop. If traffic is too heavy, I’ll use the crosswalk to get across Eastlake. It would be nice to have an official bike box to use.
@Tom. I like this as well. Sometimes avoiding the problem area entirely might be the short-term answer. I feel like doing this 3-lane maneuver daily, and the stress involved, has taken years off the end of my life. Thanks for putting it out there.
I hope it doesn’t take another unnecessary death, like Bryce Lewis in this spot years ago, as the impetus to overhaul of this intersection long-term.
I think I’d actually get rid of the bike box on SB Eastlake, and make the left turn to Fuhrman unambiguously a two-stage turn. This bike box looks kind of like the one on N 34th near the Fremont Bridge (which I have other problems with)… but on Eastlake there are really only two positions you might want to aim for: in front of the left-turn lane, to turn left onto Fuhrman, and even with the continuing bike lane, to continue straight. Between those positions is no-man’s-land in front of the two through-lanes; if the light changes while you’re moving across the bike box you might end up in a bad situation. I’d include a box for bikes turning left from Fuhrman to SB Eastlake, ahead of the crosswalk, but not a big box behind it.
All meaningless without enforcement. People who will pay attention to and adhere to this were probably already courteous and respectful drivers.
Agreed. There’s been a few articles in the Seattle Times highlighting the lack of coordination between SDOT and SPD. I’m not really sure where the blame is, either.
I agree with meanie. Strict enforcement is really the key here. After 6 months since the new 2nd Ave bike paths were installed, I still see cars turning right ignoring the red light daily. I’ve never seen them ticketed.
Elevate that section of the bike lane and ramp the sides where the cars are to cross. Extend the elevation with a curb or sign out to the end of the sidewalk (the very edge of the intersection) so that it is to sharp for a car to make a right turn from the middle lane.
This section is one of my least favorite parts of my commute to work. With driveways, parking garage entrances/exits, intersecting streets, cross walks, cars parking (illegally) and wildly changing bike lanes (lots of zigs and zags) this is a very hazardous area to navigate. I try to go as slow as I can on this hilly stretch of road and keep a very close eye on cars who are often going quite fast. Not to mention the abrupt end of the bike lane dumps you into the car lane and I’ve had to jockey for position with many cars.
I actually this this”protected bike lane” is worse than it was before. In my opinion, this shouldn’t even be called a protected bike lane as that name implies ease of use for slow/vulnerable users and it is so not easy to traverse.
“Low cost pilot designs” and unsignalized intersection treatments are not effective with protected bike lanes. This is why the protected bike lane-as-panacea mindset is so damaging: they are very expensive to implement safely, because they require full signalization (and no-turn-on-red) for every intersection.
Streets where full signalization (including signalization of currently unsignalized intersections) is too much of a burden should not have protected bike lanes installed on them, because they are inherently unsafe.
Our very limited infrastructure dollars would go a lot further focusing on implementing these in the places where they can succeed, and not wasting our money on half-implemented measures that make our streets less safe for our most vulnerable users.
The issue is that people on bikes need to use streets like Roosevelt to get anywhere in a city with geography as constructed as Seattle. There is no alternative.
I do agree with you as far as protected bike lanes go. I don’t like them. I’ve ridden on Dexter 500 or 600 times and the experience is considerably worse now than before the “improvements.” Now I have to check for turning traffic at every single intersection, since sight lines have been considerably reduced. It’s stressful. Niro mention frequent parking violations in the lanes. One day two separate police cruisers were writing tickets in the bike lanes. All four cars were parked in the lane. It was ludicrous.
Wait… in your 1st diagram the bike lane portion at the intersection wasn’t painted green, in the actual photos the bike lane is solid green. Isn’t a solid green bike lane usually reserved for “no cars” everywhere else in the city ? It seems like it should be a broken green pattern there too.
Why not put a little graphic of an “automobile” in the “right turn only” while you’re at it ?
The solid green portion in the photo looks like it’s past the patterned “mixing zone.” Solid green bike lane on the left, right turn lane on the right.
Except there’s not actually enough room for a car to make a right turn without being in the green painted zone prior to the intersection. It’s still a “real world” deviation from what was shown in the plan and is training drivers to drive over green box zones.
https://twitter.com/NEGreenways/status/592097178362871808/photo/1
I’ll have to ride through the area again to really have a good idea what it looks like (I happened to go through there while it was being painted), but I don’t think the crosswalk-like design really notifies drivers that they should merge right to turn right. The design used on Dexter probably does a better job of that, and it’s not clear why the city would use such different designs with such similar intent.
[…] fix: The Seattle Department of Transportation rolled out some small changes to the Roosevelt Way protected bikes lanes in the University District this week in response to […]
Sorry guys but that protected intersection video has been debunked pretty hard on reddit. The whole thing fails if more than 1 car wants to turn right, or if a car stops in the crosswalk (happens all the time) and makes it a big pain to turn left for cyclists. Spent last summer in Europe and in the Netherlands (only country we visited that really does anything close to this) does it considerably different. Remember that roundabouts are the norm so the whole thing works differently.
The maker of that vid is really just a bike enthusiast, not a city planner. Sorry to rain on your parade, I’m a huge supporter of better traffic flow, especially for cyclists, but I’m a bigger supporter of reality.
Edit: check out the comments from the original vimeo link and you’ll find a better explanation than my flippant claims
I don’t think the video has been “debunked.” It’s more nuanced than that.
It’s certainly shouldn’t be taken as a straight-up design guide (it doesn’t claim to be one) since all intersections will be different and have different needs. Certainly Roosevelt/42nd looks nothing like the intersection in the video.
But the concept of having a car-length space for turns so people biking and driving can better see each other is a proven concept. So is the idea of having people on bikes wait at an advanced stop line to prevent right hooks when the light turns green and the idea of a concrete island to reenforce a tighter turn radius (vital for slowing turning cars).
David Hembrow, who has been critical of adopting this design whole hog, prefers other best practices for intersections, like either A: Have little enough and calm enough traffic to not need any traffic signals (or even bike lanes) or B: Have an all-way green phase for bikes separate from cars. Neither of those options seem all that applicable at this particular Roosevelt intersection (though the low-traffic option sure sounds great): http://www.aviewfromthecyclepath.com/2014/02/the-myth-of-standard-dutch-junction.html
I like Mark Wagenbuur’s nuanced explanation of the so-called protected intersection concept. He shows a bunch of different examples around the Netherlands that use the ideas in a lot of different ways: https://bicycledutch.wordpress.com/2014/02/23/junction-design-in-the-netherlands/
The added challenge on Roosevelt is the downhill. The fact that the city is determined to maintain two lanes in the same direction doesn’t help, either. With all that extra speed, they have to figure out how to slow everyone down and provide good enough sight lines that collisions can be avoided. Separating the turning signal phase from the bike phase could help, as well, though that might be hard while also trying to maintain two through lanes (I would love to see them drop the second lane, as advocates have suggested).
Remember also that all these concepts also have to conform to US road standards. That’s one value in the NACTO guide, it takes best practices around the world and presents options that conform to US standards. If a traffic engineer veers from a standard and someone gets hurt, their job could be on the line.
That’s a long answer to your comment. But don’t be so ready to throw in the towel on these ideas. We have to solve this problem, because intersections like this are not rare in Seattle, so we have to make them safe.
Any talk about turning Roosevelt and 11th into 2-way streets again? That would probably make them more bike & ped friendly.
For supporting evidence, take a look at Mercer between QA ave and 5th N. It still handles a huge amount of traffic but is much more accommodating to peds. I can’t say whether it’s better for bikes because I’ve been using Roy (which has practically no traffic).
@Eric.br – this small intersection got so much attention because I work at that corner, and along with my coworkers witnessed many collisions. We started calling 911 every time a car hit a bicylist (because after awhile we realized the shocked cyclists just wanted to be on their way without realizing the city needs a paper trail), and we all wrote to the city. Our fear was that one way we’d witness a death, that’s how many accidents we saw out our window.
I don’t know if this new design is perfect, but since its implementation we haven’t seen an accident. Happy the city took notice!
Lorraine, thanks for reporting those collisions! I’m curious if you’ve witnessed any more since the changes were done a month ago.
Looking at the SPD collision stats, there were 3 (reported) right-hooks in ALL of 2014 at this intersection, and 2 reported in just the first 4 months of 2015 (after the temporary PBL installation). Looking at twitter reports, there were clearly many more that occurred. Given that, I’m reluctant to rely on just SPD data.