Bike to Work 4: Best of all worlds, together

fahrradstrasse

In a previous post, I pointed out that the popular “protected” bike lanes in fact are less safe than cycling on the street: The protection ends where cyclists need it most – at intersections. Here I propose an alternative model for getting more people to cycle without increasing the accident risk. It relies on appropriate facilities for each situation, depending on traffic speeds, intersection density and other factors. The example of Munich in Germany shows that this approach can work well.

Different facilities for different situations
Data from the Belgian city of Antwerp looks at the relative accident risk (with the average risk being 1) on streets based on the speed limits and bicycle facilities:

table_antwerp

The speed limits really are just a way of distinguishing different types of streets: Neighborhood streets with low speed limits (12/20 mph), busy city streets with many intersections (30 mph) and major roadways with few intersections (45 mph).

The biggest accident risk occurs in separated bike lanes that run along low-traffic neighborhood streets with many intersections. In this case, riding in the bike lane is similar to riding on the sidewalk, which is known to be the least safe location for cycling.

Generally, where cars are moving slowly, it’s safest to ride on the street. However, on major roadways with high speeds (and few intersections), separate bike lanes are safest.

In between these extremes are streets that have a lot of traffic and many intersections. Here, the best approach lies somewhere between no infrastructure and a painted bike lane. These streets still display above-average safety, but they are not quite as safe as riding on the roadway of low-speed streets or on separate paths along highways with few intersections.

I think this statistic can point the way forward with a multi-pronged approach for bicycle facilities:

no_path

No infrastructure at all

  • On streets where cars and bicycles move at similar speeds.
  • Speed limits must be 20 mph or lower.
  • Best for streets with multiple intersections, as it keeps the cyclist in the visible part of the roadway.
  • Streets like these also are great candidates for “bicycle boulevards,” which give the right-of-way to cyclists (see below).

Bicycle_Lane_1

Painted bike lane

  • On streets where the speed differential between cars and bicycles is relatively small.
  • Speed limit must be below 30 mph.
  • Suitable for streets with a moderate number of intersections, as it keeps the cyclist close to the main flow of traffic, hence visible.

motorwayjensdressling

Separate path

  • On streets where cars move much faster than bicycles.
  • Suitable only for roads with very few intersections, since intersections present a great danger with this design.

Appropriate Design
It is important to realize that there are three parameters that are inter-related:

  1. number of intersections
  2. speed limit
  3. bike facility design

The number of intersections cannot be changed (unless you build bridges and underpasses). It will dictate the design of the bike facility in most cases. The other two parameters, speed limit and bike facility design, should be adjusted to find the best solution for each roadway.

For example, if the traffic speed is too high for a painted bike lane in the roadway, but there are too many intersections for a separate path, then the speed limit should be decreased to make the painted bike lane safe. Fortunately, traffic speeds have to decrease as the intersection density increases, so the needs of cyclists trend in the same direction as those of other traffic. Decreasing traffic speeds a little further will make things safer for all.

With these simple guidelines, it should be possible to design bike facilities that both are safe and feel safe to cyclists. The example of Munich (Germany) shows that it is possible to make cyclists feel safe and get more people to cycle, without pushing them off the street onto separated cycle paths.

fahrradstrasse

Munich and Bicycle Boulevards
Munich, southern Germany’s biggest city, has embarked in recent years on an ambitious program of improving its bicycle facilities. The city has been removing or modifying facilities that have been shown to be unsafe. New facilities are either constructed as painted bike lanes or “Fahrradstraßen” (“bicycle boulevards”, above), which are turned over to cyclists as the main users. Cars are still allowed, but considered secondary users.

Hinweisschild_Benutzungspflicht_neu_d1f692e79f

Munich also is installing signs that legitimize cycling on the street (above), even where there are separate paths. This is a big deal in Germany, the country that started the trend toward clearing the road of cyclists to provide unobstructed room for cars.

Combined, these measures have increased cycle use in Munich (as a proportion of all trips) by 70% in 9 years. 17.4% of all trips in Munich are made by bicycle. This growth far exceeds that of those German cities where segregated cycle paths remain the preferred option.

It is well known that better facilities, where cyclists feel safe, bring out more people on bikes. Munich shows that this goal can be achieved without segregated trails and without compromising safety at every intersection.

Whereas European cities often do not lend themselves to the creation of “bicycle boulevards,” North American cities provide an ideal setting for streets that are dedicated to bicycles. Here is why:

copenhagen

European cities usually have an “organic” street layout that grew over time (Copenhagen shown above). For any given destination, there usually is only one relatively direct route. This means that cars and cyclists have to be routed on the same, busy streets. Creating a “bicycle boulevard” means closing an important traffic route for cars.

seattle

Most North American cities are built on a grid (Seattle shown above), which provides multiple routes for each destination. This makes it easy to provide separate “arterials” for bikes and cars. (The map above shows the “car arterials” in yellow.) Creating “bicycle boulevards” is relatively easy in this setting.

Bicycle boulevards are the ultimate separation, because they channel traffic flow differently for cars and bicycles. Even where bicycle boulevards cross “car arterials”, they do so at right angles, which eliminates the “right hook” and “left turn” hazards. These hazards occur when cyclists and turning cars “share” the same intersection. Where a bicycle boulevard cross a car arterial, either the cars or the cyclists have a green light, so they do not “share” the intersection.

out_of_town_seattle

Mark, one of Bicycle Quarterly’s bike testers, pointed out to me that we privately already use what effectively are bicycle boulevards. We have mapped the city to find routes that are fast, efficient and yet away from the main traffic arteries as much as possible (above our northward route). Formalizing these routes as “bicycle boulevards” would be a win-win for all cyclists, and even local residents, who’d enjoy quieter streets and increased safety from cyclists keeping an eye on things.

One thing that is important for efficient bicycle travel is to give bicycle boulevards the right-of-way over cross-streets. Otherwise, they are little different from neighborhood streets with a relatively high accident risk at every intersection.

All we need in North America is the political will to dedicate a few streets – which see little traffic anyhow – as bicycle boulevards. Give them the right-of-way over cross streets, signal the intersections with car arterials, and install appropriate traffic calming devices to sure that cars don’t use them to avoid congestion elsewhere. It’s a relatively simple solution, and we are lucky to have the conditions in place that makes this possible. We could be the envy of Copenhagen cyclists, as they ride on narrow, congested paths and have to deal with turning traffic at every intersection.

Of course, bicycle boulevards cannot go everywhere, so there will still be a need for on-street bike lanes. And separate trails can provide good options in some places. This basic guideline also needs to be adjusted for local conditions.

Conclusion: Better facilities bring out more cyclists. Which facility is best and safest depends on the conditions. North American cities are uniquely suitable for bicycle boulevards, which provide the ultimate in “protection” and “separation” by channeling car and bicycle traffic on separate, parallel routes.

Postscript: For those living in Seattle, the new bicycle master plan, which includes many miles of segregated cyclepaths will be presented at open houses in June. This is an opportunity to make your voices heard. Click here for dates and locations.

Click here to read more posts about cycling safety, cyclepaths and bike lanes.

Photo credit (path along freeway): copenhagenize.com

About Jan Heine, Editor, Bicycle Quarterly

I love cycling and bicycles, especially those that take us off the beaten path. I edit Bicycle Quarterly magazine, and occasionally write for other publications. One of our companies, Bicycle Quarterly Press publishes cycling books, while Compass Bicycles Ltd. makes and distributes high-quality bicycle components for real-world riders.
This entry was posted in Cycling Safety. Bookmark the permalink.

62 Responses to Bike to Work 4: Best of all worlds, together

  1. cbratina says:

    Jan, I really appreciate your studying and reporting on bicycle safety. I have found too many people who want to stick bikes onto bike paths and then forget them, and bike path maintenance. I whole heartedly agreed from my experience and am passing your commentaries to community activists working on bike safety. THANKS.

    • Sam says:

      I’d like everyone to know that Berkeley, CA instituted a Bicycle Boulevard program essentially exactly as Jan describes here and it is a wonderful success. The smaller streets are labeled as BBs, and given some right of way priority. It makes it easy to get everywhere. The only problem is we still need to cross some very busy streets, but there’s no way around that.

  2. Jan,

    I was particularly interested in the chart showing risk. I wonder if there is a correction that takes into account the types of cyclists using each type of facility. For example, by far the highest risk is for the separated bike lane installed on the low-speed street, exactly where children and inexperienced adults would choose to ride.

    Bob Cooper

    • I don’t know whether cyclists are required by law in Belgium to use separate bike lanes where they exist, but in practice, almost all cyclists do. So it’s not that you have experienced cyclists riding on the road and inexperienced ones riding on the path, side-by-side.

      There could be a selection due to route selection, where experienced cyclists avoid routes with separate bike paths, whereas inexperienced cyclists prefer them, but I don’t think it would skew the results so much that it would change the ranking. And such a selection, if it occurs, would reinforce the point that cyclists who act based on real, rather than perceived, risk, prefer to ride in the street.

      • Erik says:

        (Since I gave the data, I’ll answer :-) ) The full paper can be found here: http://www.verkeerskunde.be/sites/default/files/675_eindrapport_fietsongevallen_pz-antwerpen__16_september_2011.pdf.

        The table I referred to is on page 21. It’s title reads: “Accident density related to bicycle infrastructure and speedlimit”.
        Your question implies that certain infrastructure is used more by a certain type of rider, this is only very limited the case in Belgium. Our road network is very dense and when making trips it is almost impossible not to (al least) cross a very busy road. On page 9 you see a graph which shows the occurence of cyclist in accidents based on age. The biggest group is the 25-59 group, which are the commuters.

        @Jan: If there is a sign indicating cycle path, then it is obligatory. However if the rider desides the path is not rideable or holds a risk (snow, bad state, not suitable for the riding speed,…), it is always allowed to use the road. Obviously there has to be some ground of truth in that, mostly all riders use the cycle path if available. Only larger groups of “racers” use the road. It is important however to remember that junctions, certainly important ones are more and more designed to Dutch standards (and conflict-free). I tend to disagree with Jan on the need for cycle paths (I would focus on better design), but I fully agree that creating cyclepaths without thinking about the junctions is bound to give dangerous situations, especially in an area where bicycles are a minority and cardrivers are not used to giving way to cyclers. (In Antwerp around 20% of all movements are by bicycle).

      • Cyclepaths appeal to many cyclists. I don’t think it’s a question of whether they will be built – they will – but how to build them and maintain or even improve safety. Many current designs compromise safety.

      • Erik,

        Sorry I didn’t give you credit, but I only had your first name from the comments…

        Jan

  3. Andy says:

    I believe advocates are trying to replace the term “Bicycle Boulevard” with “Neighborhood Greenways.” These are likely not boulevards (wide streets) and they are not only for cyclists – it’s more about opening the streets to all active transportation as a priority over motor vehicles.

    While the greenways have the potential to work well, they too can be implemented poorly. There are/were plans for them to be used in Ithaca, NY, but the plans degraded horribly over time. The initial draft showed both east/west and north/south routes, making a nice grid across the city of medium density roads. The drafts slowly took out the straight paths, and replaced them with winding low traffic routes. This essentially kills the purpose of a Neighborhood Greenway.

    The point is not to find the lowest traffic route and put up a few signs hoping that cyclists will go out of their way to use it. Many cities tried this with green “Bicycle Route” signs long ago, only to find that people simply don’t care to go out of their way to reach a destination. Instead, the best greenways seem to be the ones that are one street away from the busier roads, while still being an effectively straight and nearly direct path to popular destinations. These streets often have too many lights and stops signs along the way through, so a good greenway design would flip the signs to make it more direct for cycling, by having the low traffic cross street traffic stop instead.

    I think in the search for perfect, the planners have created awful. I already have the ability to ride low traffic streets with a turn at every block if I want to avoid cars, but it doesn’t get me anywhere. We already have great recreation paths and many more in the works in the next few years; people looking for a <10mph bike ride have many options. But those of us looking more for a 15-20mph commute route are just stuck between the main through streets by mixing with traffic, or taking the side streets and planning on hitting a red light nearly every block.

    • I agree, and that is why I use the term “Boulevard.” It conjures up a pleasant route that sees a lot of (bicycle) traffic and that has the right-of-way over cross streets. Unfortunately, the last (and important) part doesn’t seem to be part of the plan even in Seattle, from what I can tell. Which in turn makes the whole concept next to useless.

      • marmotte27 says:

        Yes, one always comes back to what I deem to be the central point which must be met BEFORE anything is done in terms of infrastructure. There has to be the will and the decision to give space and priority to bikes and therefore TO TAKE IT AWAY from cars!

        I’m glad to here that it seems to be the case in Munich, the increase in the modal share of cycling seems to confirm it. It stagnated around 9% for decades in Germany because they really wanted to relegate it to the sidelines rather than favour it.

    • I agree that good intentions often lead to cycling hell. Perhaps more effective would be another approach that the City of Albuquerque used some 5-10 years ago: stripes on certain cross-town streets chosen, *not* only because they were less trafficked, but because they were strategically placed for real cycling commuting, this combined with sharrows, signs reading “Share the road”, and a TV (and radio?) ad campaign with the theme, “Share the road.” I must say that, in my own experience of commuter W-E across town on busy, rush-hour, arteries, these had, at least, the subjective effect of making me feel more secure — no more “Get off the road!” responses from motorists. Once again, this points to the principal remedy: cultural acceptance of cyclists as a banal, quotidien part of the morning and evening commute.

  4. I generally agree with your taxonomy, with one qualification: I think striped bike lanes even on fast, multi-lane boulevards, are a plus for cyclists, for two reasons: first, they help curtail — or, “perhaps” they help curtail — motorists’ annoying penchant for drifting to the far right of their lanes. I know this happens even on striped streets, but I am guessing that a striped bike lane provides a leftward alternative to the curb for such drifters. (OTOH, I notice that, even on striped bike lanes, I get more clearance from cars when I ride toward the left of the bike lane. Thus I think one great help would be to have police enforce the rule that autos must stay out of bike lanes. Perhaps this sort of rightward drifting is a motorists’ penchant unique to the US, or even to the SW US? The very, very odd opposite is the tendency of other, more conscientious motorists, to fearully scrub their tires on the left-lane curb when passing cyclists.)

    Second: The bike lane is a visual reminder that cyclists may be present, and — perhaps this is a third reason — it strengthens the perception of cyclists’ legal right to be on the road.

    Once again, in the face of all these alternatives to make cycling safer, I strongly believe that the real remedy is cultural, that is, just a sufficient awareness among motorists that cyclists legitimately use the road. Compared to this, all the infrastructure in the world is in second place.

    Generally speaking, I think that the right way to discern what cycling infrastructure works is to observe in depth what cyclists and motorists do already. Cyclists seek quieter streets, but they choose those streets more populated with cyclists. Motorists become more accommodating when they get used to driving among relatively large numbers of cyclists.

    • I agree that bike lanes are better than nothing on fast, multi-lane roadways. But I also think that those streets – assuming they have few intersections – are prime candidates for separated bike paths. And if there are too many intersections for separation to be implemented safely, then the speed limit should be reduced.

      • We have to distinguish between ideal and workable. In my riding environment, no power, either political or cultural, is going to reduce the 40 mph posted, 50-to-60 mph actual speed across, for example, Montano Boulevard (ABQ, NM) to 30-35 mph. While the sufficiently wide lanes are the all-in -all for cyclists using this route, the striped bike lanes do alert motorists to the presence, legal as well as actual, of cyclists. The lines are less of a physical barrier than a cultural and legal reminder.

        Note that I write from ABQ, NM, and that the example here — Montano Boulevard — is one of the few crossings of the Rio Grand that separates westside Albuquerque from eastside Albuquerque. These bridges are always packed at commuter periods. No one is going to persuade the city council to reduce speed limits on these arteries. But the particular cases have a lesson for the general rule.

      • We need to first identify the ideal, then ask for what is feasible, and finally accept what is politically workable. In that order.

      • Andy says:

        Now I’m lost. Your chart above shows that painted bike lanes are always less safer versus no paint/separation. In fact, that chart indicates that the only time that bike specific lanes are safer is on 45mph roads with a separated lane. So why are you now saying “I agree that bike lanes are better than nothing.” ??

      • I don’t think the small differences in the chart are statistically significant. And they don’t take into consideration the context. From my own experience of riding on multi-lane roadways with high speeds and lots of traffic, I vastly prefer a bike lane over nothing at all.

  5. Mike Beck says:

    Tiered bike facilities with different classes of facility his all sounds great in theory, and one shared by most transportation planners tasked with moving more people by bike in european and north american cities alike. The difficultly for any city lies in developing appropriate, context sensitive designs with considerate intersection treatments, including a reiterative feedback loop to assess safety and work to improve what works and what doesn’t.

    The methods differ. Some feel the German designs substandard, and more akin to the sub-par US ones, rather than world class cycle track and separated path networks seen in well facilitated regions in the Netherlands.

    One thing to point out to Jan is there are a lot of American streets that don’t meet these delineations. There are a lot of streets and arterial collector roads where a simple bike lane ISN’T enough. Most of the roads in Bellevue, for example. These roads merit some degree of buffering from the car traffic, if not greater separation somehow on the select main routes.

    Jans’ plan overlooks the entire panopoly of buffered and floating parking/intersection setback designs in use in america. on some roads, a buffered bike lane is going to be more cost effective than plopping a separated path down.

    Market Street in San Francisco has become improved as it beefs up its separation from car traffic in spots where, depending on context, it was needed. For example, adding bollards at a spot motorists made illegal turns to enter the freeway.

    There are plenty of arterial and collector roads – the direct routes cyclists would like to take, along which a more robust separated or buffered facility as one of the route choices makes sense.

    For a specific example – how do you route kids on bikes from the Seattle Zoo to the Seattle Center?

    This seems like a simple enough task, but i’m not sure bike boulevards or bike lanes are going to be appropriate on all of the roads in this case. Seattlites would be well served with a more established, protected bike route from north of the ship canal leading to downtown from Fremont. Protected bike lanes on part of the route is almost a given for this specific yet crucial transportational route into and out of downtown Seattle.

    And given political will, there’s a way for seattle to do it (Hint: Westlake cycletrack, aurora underpass separated track)

    Same thing with the missing link. completing the missing link and getting rid of one of the most unplapable choke points on the Burke Gilman trail/ “Cycletrack” is a desirable goal. Planning how to manage the superstore parking lot crossings, the cement trucks, and routing cyclists on the wrong side of the street at the intersection of 28th and Market is worth the effort in designing smarter facilities and better attention to intersection treatments.

    I like Jan’s plan for using tiered levels of context sensitive bikeway designs, to best serve all classes of rider. Consideration should be given to all riders to feel safe, and to be as safe as design allows, when choosing to get across town on a bicycle. This planning dictate is shared in common in bike planning the world over. The difficultly is in the execution.

  6. Harald says:

    Do you have a source for the Antwerp data in the table? I’m curious about their methodology and how they were able to reach statistical significance with the generally low number of fatalities.

    I would guess that at least the small differences in the 30mph category (I guess that is actually 50km/h?) are not statistically significant, from which one could draw the conclusion one should build bike facilities on all those streets, as they strongly increase subjective safety (and therefore ridership) while at the same time not being less objectively safe.

    • Erik says:

      A report with the data can be found at: http://www.verkeerskunde.be/sites/default/files/675_eindrapport_fietsongevallen_pz-antwerpen__16_september_2011.pdf. I’ll translate the relevant parts to answer your questions:

      -Antwerp keeps “enriched data” of the accidents that occur within their jurisdiction. This means they also keep track on roadfunction, cycle infrastructure, priority, speedlimit, speedreducing infrastructure, traffic intensity,… (normal data only contain things like date, location, people involved, …)

      -The research is based upon a database built between 2000 and 2010, containing 83588 accidents, 8184 of which with bicycles involved. They state that probably only a part of the accidents are reported (for small accidents people tent to come to an agreement without police involvement).

      -The 30 mph category is indeed the 50 km/h limit. It is difficult to draw direct conclusions from these data since it is unknown how much cycle traffic there is in every street. However, the researchers state that the infrastructure should be matched to the environment: in a residential area, cars drive relatively slow and expect to see childeren and bicycles on the roadway. They do not expect a cyclist to pop up at an intersection from a hidden cycle path. Likewise on a boulevard with two or more lanes and often heavy traffic, they do not expect bicycles on the roadway, but are very aware of the interaction with pedestrians/cyclist at intersections.

      If you read the paper (assuming you can translate it) you’ll notice the number of accidents is not declining very rapidly. However the amount of cycling-kilometers seems to be increasing every year. The recently started bicycle-sharing program registers 5600 rides a day (an average of 6 rides per available bike – in 2012 there were 2.062.359 registred rides). Measurements on the amount of cylists a day are not publicly available (if there are) but my personal impression is that there is a significant growth so the number of accidents in relative numbers is probably decreasing.

      • Data based on more than 8000 accidents in a single city at least is a bit more significant than the Havard study, which had data from 55 accidents spread over many cities.

      • Harald says:

        Thanks, Eric! Between speaking German and using Google Translate I think I understood most of the report. The fact that they don’t have traffic volumes for the respective street types seems to make the data completely meaningless, as it could merely mean that more cyclists use roads with bike infrastructure and therefore more accidents happen there. The authors themselves point out that the low number of accidents on 90km/h roads with no bike infrastructure probably just mean that almost nobody cycles there.

        And whereas they’ve done significance testing for some of the data, they at least don’t provide p values for this particular table.

        I don’t mean to question your general argument, Jan, but since you’re always quick to point out other researchers’ flawed methodologies (rightfully so!), I would caution against taking the table in the article as the basis of your argument. I mean this is one of the big problems of traffic design and engineering: getting objective, reliable, and valid data is just very difficult and expensive.

      • I agree on the low accident rates on high-speed roads. However, with 8000 accidents, and knowing how European cities are laid out, I think the data for the lower-speed streets is going to be more or less representative. You have to base design guidelines on something, and other data does support the idea that low-speed streets with high intersection density become less safe if you separate cyclists from the rest of traffic.

  7. A good place to ride is on a four-lane boulevard with narrow lanes and no shoulder. An example near my home: 43.117964,-77.555176 By taking the lane, the whole right lane becomes a “bike lane.”

  8. Greg says:

    Well-said, Mike.
    Jan, one comment that I would add to this excellent discussion is that (in my State, anyways) the law states that speed limits on local streets must follow the ’85th Percentile’ Rule. That is, open the street to traffic without limit postings, measure the speeds that motorists actually travel over some pre-determined time period, and set the limit at the nearest 5-mph increment to the 85th Percentile speed. We have recently seen several local streets’ limits increased following successful lawsuits that followed speed limit lowering events. You can’t easily set lower limits on (our) local roads, fwiw….

    • cyclotourist says:

      Could/should non-motorized traffic be included in the census? If so, it would greatly lower the 85th percentile speed.

      • Joseph E says:

        No, it wouldn’t. If bikes were even 10% of traffic on the street, this would only shift the 85th percentile level down to the current 75th percentile speed. That would often be only a couple miles per hour slower, not enough to change the speed limit (which is usually set in 5 to 10 mph increments – 25, 35, 45…).

        The law needs to change to set speed limits based on safety, not based purely on road design, or the road needs to be re-designed so that the majority of drivers slow down.

  9. tomiot says:

    Very nice write up. I very much agree with this. Ever since the bike lane was built up on 65th – Sandpoint, I’ve thought a lot about this. It’s a slow street and the separation with the barrier obstructs the car’s view for cyclists, especially at the Sandpoint intersection where cars are turning right. Made me realize that separate lanes are not always the answer.

  10. Pete says:

    Great series of posts, Jan. Thanks for the thoughtful insights. The only thing I might change is using “measured average speed” instead of speed limit. In most of America the posted limit is irrelevant to the actual speeds. The tiered approach would be most successful if it were applied to the reality rather than the ideal.

    • You can design roads so that drivers don’t exceed the speed that you want them to maintain. We often accept that cars will go too fast, but that isn’t a given. Enforcement can do the rest. (Look at France before and after they installed automatic speed cameras in most towns. Everybody now obeys the speed limit, whereas before, it was like a Formula 1 race.)

  11. rory says:

    There’s so many interesting thoughts here to promote and improve cycling in a city environment. I think the tiered system is interesting, since it matches exactly with roadway design for cars. all road design is tiered, and based on the supposed end user, whether its a local road, collector, or a major arterial. I’d also like to point out that intersection and roadway design are 2 completely separate animals due to their complexity. The thing I find interesting about this blog post is that it appears to contradict the previous blogpost about bicycle segregation by promoting bicycle boulevards.

    Bike lanes/ boulevards or separate facilities in general(HOV lanes, Parking Lanes, Etc…) are always an interesting political talking point. Not to single cycling out, but the use and dedication of public right of way to one specific mode of transportation just doesn’t make sense, unless it improves all other users in some respect. In general, roads are built to support the transportation of people, goods, and services. This includes, but is not limited to, corvettes, buses, cement trucks, unicycles, WB-67s, and a host of other pneumatic tire rolling transportation devices. oh, and occaisonally pedestrian walkers in the event there is no shoulder/sidewalk. When we start taking public right of way and saying it has a dedicated end user, I would expect that faciltiy to have the numbers or the political interest to accomplish this.

    I may not understand the end design of the “bicycle boulevard”, but it still sounds like segregation of bicyclist from cars(regardless if it’s actually a shared facility). Any form of propaganda that promotes a bicycle exclusive street is going to be met with the same hurdles as a bike lane. I enjoy my neighborhood in Seattle, and seeing riders traveling through it. They do so because they are practicing their own judgement about traffic conditions and safety(like you and Mark). I don’t think dedicating 44th as a greenway or bicycle boulevard has honestly made any difference whatsoever, but I hope I can be proven wrong and we’ll see a report on it’s performance.

    I also don’t understand how bIcycle boulevards eliminates conflicts at an intersection. Intersections have conflicts and turning movements. that is the performance charactaristic of an intersection. Trying to make an intersection appear safe and conflict free seems futile. I’m trying to understand the right angle comment, since that would either require a right turn (few conflicts) or a left turn (a lot of conflicts). Also, regarding traffic signals and adding a phase to an intersection for bicycles doesn’t make sense at these particular low car volume neighborhood intersections. if the dominant movement is the cyclist, then why is there a need for separate phases? If there is a car, they can wait in queue with the cyclist. Either way, what seems to be proposed also contradicts the idea that there have “no infrastructure at all”.

    Also, dedicating a street as a “bicycle boulevard” does not necessarily increase safety, and will not necessarily make residents happy. by dedicating the street, the hope is to increase the volume of user to that street. if the roads have any grade at all, and are a local street, i would be concerned about a fast traveling cyclist hitting a kid coming out from a driveway. While rare, the likelyhood of this occurence just went up due to wanting to increase the volume of cyclist on the road.

    As an alternative to “Bicycle Boulevards”, I would suggest that more effort be put to show these routes through grass routes cycling campaigns hosted through local organizations (I’ll begrudgingly say cascade) and social media. The route Jan just showed just got way more exposure then the city putting a map out and saying it’s a Bicycle Boulevard. I would rather see a service that can guide riders through the neighborhood, rather then showing it on a map, which has it’s own inherent flaws (this local e-w road is better then this local e-w road 1 block away).

    • Separation is great – I love riding on roads in the mountains that don’t have cars, because they are closed for motorized traffic. But it must be real, full-time separation, not part-time separation that puts the cyclist back into the mix at every intersection.

      If you have two overlaid networks of streets, one preferentially for cars, and one preferentially for bikes, you also make the intersections less complicated. When bikes and cars go in the same or opposing directions, you have conflict at intersections if one user wants to turn while the other wants to go straight (right hook, left turn hazard). If bikes and cars meet only where a Bike Boulevard crosses a Car Arterial, then they are going at 90° angles of each other, and there is no conflict at the intersection. Either cars or bikes have a green light…

      I agree with you that education for route finding is something where money might be well-spend. Imagine if instead of spending millions on a “cyclepath,” the city provided a “route planning program” where you can e-mail your destination and get a good cycling route – sort of like the bus planner that is online! Instead, novice cyclists think they need to ride on the streets where they drive, and they find that they don’t like it. If we showed them the “secret” routes through town, they might see how easy it can be.

      And if we then make these routes even better by eliminating the intersection danger at unsigned intersections, we would be a long way toward a great cycle-able city.

      • rory says:

        This is where I guess I’m confused. I understand a bike boulevard to be a shared facility, where bicycles have the priority. as the bike(shared) boulevard comes to an intersection, there are still cars/trucks wanting to turn onto or off of that particular boulevard. If that access is eliminated, I have a tough time thinking the residents that live on that street are going to be glad the road they live on is a bike boulevard.

        regarding separation, there are formulas that show a pedestrian/bicycle level of comfort. this level of comfort goes up as the adjacent street’s vehicular volume/speed goes down, as well as the separation becomes greater. Yes, I agree that eliminating points of conflict are great, but this becomes very difficult in an urban environment, outside of grade separation.

      • The idea is that a Bike Boulevard doesn’t have enough cars to make them an issue. Basically, cars and trucks are allowed only to travel a block or two to reach individual houses, but not to use this as a thoroughfare. And since bikes clearly have the right-of-way on the Bike Boulevard, the cars will not expect to push past them…

        If a Bicycle Boulevard just means putting up a few signs without changing the right-of-way and access to the street, then it’s useless.

    • Robert Cooper says:

      Today Dan Gutierrez, writing on the “Cyclists are Drivers!” Facebook page said: “The engineer created a theoretical deign for others to use. This is what happens when designers lack basic knowledge.”

      I posted this reply: I live near the center of an urban-suburban area of about 700,000 population. (Traffic.) Yesterday, I cycled to visit a friend about fourteen miles away. I left from my driveway and did a loop that included his house and that totaled 55 miles. (Not content with only 28 miles for the day.) I may have seen some cycling-specific infrastructure, but I didn’t notice. Conflicts with motor traffic: Zero.

  12. GAJett says:

    In May 2013 “Harvard School of Public Health (HSPH) researchers … say that U.S. guidelines should be expanded to offer cyclists more riding options and call for endorsing cycle tracks — physically separated, bicycle-exclusive paths adjacent to sidewalks — to encourage more people of all ages to ride bicycles.” HSPH then calls on the American Association of State Highway and Transportation Officials (AASHTO) to update its guidelines to provide more separated paths. “http://www.sciencedaily.com/releases/2013/05/130516161657.htm”, and “http://ajph.aphapublications.org/doi/pdf/10.2105/AJPH.2012.301043″.

    While the Harvard research provides some statistics re. crashes on bike paths and physically separated bike lanes, and refers to European data, it seems they ignore the data that Jan mentions above and perhaps his very sensible solutions.

    My biggest concern is, if the Harvard recommendations are adapted by AASHTO, there will be a big push to create separated bike lanes. However, in this current economic climate there is little money to do so and this will become an excuse to halt the placement of cycling paths of any kind along roadways (too expensive for separated facilities, too dangerous for painted separators).

    • The Havard study’s crash statistics are deeply flawed, as we discussed in an earlier post. The biggest problem is that the study compares accident rates on cycle tracks with the general accident rate in the U.S. That is highly misleading, because we all know that cycling is safest where more people cycle. Cycle tracks also exist where most people cycle. So of course, a cycle track in Seattle will be safer than riding on the roadway in rural Florida! In the very least, they should have estimated the on-the-road accident rate for bicycles in each municipality where people cycle.

      I was surprised that the study made it through a peer review process with such glaring problems. (They also didn’t bother to examine whether their results were statistically significant.)

      Unfortunately, I have seen the Havard study quoted all over the place as “proof” that separate cyclepaths always are safer, when most data indicates that installing cycletracks increases accident risk. My hope is that this discussion will help to introduce the concept of “intersection density” to the planning vocabulary.

      • I’m aware of at least one rebuttal (in development) to the Harvard study – their own data, when sorted for intersection density, show that intersection-dense urban cycletracks are more hazardous than roads without infrastructure.

        Only by conflating a large number of suburban/low-intersection-density sidepaths with a small number of urban/high-intersection-density cycletracks were they able to engineer their conclusion that cycletracks are safer. (In Seattle terms, they’re using the accident rate of the I-90 Bridge trail to estimate the accident rate on Westlake.)

        Like you, I’m somewhat surprised the study made it past peer review. I suspect that’s because it was not peer reviewed by people who ordinarily study transportation engineering, so what seem like obvious flaws were not obvious to the reviewers. “Intersection Density” is *already* part of the transportation planning vocabulary, but it may well be alien to public health reviewers.

      • Even if the reviewers don’t study transportation engineering, the lack of statistical analysis should have raised immediate red flags. And the comparison of data from very select cycletrack locations with a national accident rate also should be obvious to anybody with scientific training.

  13. Mike Beck says:

    I find it interesting in Jan’s typology, that roads with speed limits of 30 miles an hour and above merit separated facilities. bike lanes are only suitable for roads with traffic speeds below 30mph.

    In a sense, Jan is looking at a city like Seattle and suggesting that as the road designs stand now, there needs to be a lot more cycle tracks. Certainly, the ones along the docks and piers he’s endorses cycle tracks on aren’t only roads in Jan’s typology that meet his separated lane criteria.

    Oh, and Jan -the phrase bike boulevards is already dated and passe. it’s “neighborhood greenways’ now, because biking is going to continue to be seen as a special interest in americans’ minds for some time to come.

    • Mike,

      You misunderstand. Intersection density is the first variable, since it is difficult to change. So if you have a lot of intersections and a speed limit above 30 mph along a major bike route, then the conclusion is not to put in a segregated cycletrack, but to reduce the speed limit.

      “Neighborhood Greenways” appear to be a watering down of the concept of bicycle thoroughfares where cars are tolerated, and where bicycles have the right-of-way over cross-streets. They seem to amount to little more than just a few signs posted inconspicuously to avoid annoying anybody. We need to demand better than that.

      • Bob says:

        If you’re going to offer a critique of Neighborhood Greenways, it would be best to first become familiar with the goals of the group and the individuals advocating for those goals. We are very aware that intersections are where the greatest danger lies and where the costliest improvements ($$) are required. Every intersection requires individual planning, but many times we advocate for traffic diverters on arterials so that bikes can cross but cars must turn, which reduces cut through traffic. Unfortunately, prioritizing even a couple intersections for bikes and pedestrians over cars is met with fierce opposition from powerful interests. The 58th St Greenway in Ballard was originally designed with a more aggressive diverter for 24th Ave, for example, but the apartment building managers twisted SDOTs arm so that cars would not be inconvenienced.

        Far from being a ‘watered down’ version of bicycle boulevards, it expands on the concept to include cycling and walking for users of all ages and abilities. I think if you learned more about the group and its grassroots origins you would find much in common with them. We could use your help and voice in advocating for more aggressive pro-bike treatments at key intersections in the city.

        http://seattlegreenways.org/

      • I really appreciate your work. I am glad you are working toward giving bicycles the priority on these streets. As I have worked in cycling advocacy myself, I understand the forces that you are up against…

  14. Nathan Backous says:

    It’s a shame that politicians/”advocacy” groups like those in my hometown of Portland Oregon are seemingly obsessed with the idea of “world class” (can we please cut that phrase from the style guide?) separated bike paths and trying to jam them on intersection dense roads. Also, they seem to be obsessed with “multi modality” and try to fit street-car tracks, auto lanes and bike lanes all onto the same pavement. An expensive disaster just waiting.

    Sure those plans look great in artists renderings but it seems to me that none of those involved with the planning have the slightest idea about what it’s actually like to ride a bike (or drive a car for that matter) in an urban area.

  15. Edwin W says:

    When I lived in Berkeley, I liked their Bicycle Boulevards. The main, big, arterial streets were (mostly) for cars, and the residential streets were made to make it hard to drive through them – lots of forced turns, dead ends, etc… where cyclists (and emergency vehicles) could continue straight. When I first moved there, it was a drag to figure out how to drive to certain locations, but once I figured it out, I lived with it, even if it meant driving a few blocks out of my way. I think the residents on those streets did not mind driving a few blocks out of their way, either, as their streets were so much quieter, with the only through traffic was bikes and walkers.

  16. Stephen says:

    Very well said Jan. Unfortunately, here in Sydney our street layout is generally “organic,” or else it follows the topography in hilly areas, so the scope for parallel routes isn’t so high. Our other problem in Oz is that things here tend to lag 5-10 years behind what happens in Europe or North America, so I expect we will be repeating some of the known mistakes for a while yet; the rise of the internet will hopefully reduce this lag over time.

  17. Heather says:

    Great points. Vancouver, Canada has a bicycle boulevard type system which I always found to work fine. Major intersections have crosswalk lights which you can press, and are given the right of way at small intersections. The only concern I have is that cars still used them too much, locals parked their cars on either side, so in practice there can often be little room left for bicycles and cars. The separated paths set up downtown are good because they get cyclists out of the way of confusing one way streets and super impatient traffic. The separated paths are a fairly good way to get around downtown and then onto the bike boulevards towards a destination. But intersections can be an issue.
    A cyclist was just killed in Vancouver when she bounced off a sidewalk on a busy causeway. Cyclists must use sidewalks through the causeway in Stanley Park, and is a major artery for cycling commuters, often in a rush. Cyclists must give way to pedestrians on the sidewalk whom would be better off sticking to trails through the park! At any rate, pedestrians standing around aimlessly is a recipe for disaster as the woman was trying to avoid some people on the sidewalk and fell off, hit by bus. So, now talk of a barrier or separated lane. It is in fact far more pleasant to get off the causeway and ride through Stanley Park into or out of Vancouver. I am scared of cycling on that bumpy sidewalk on the causeway, so always go through the park instead. As I lived in a city with no infrastructure for years, I simply mapped out my routes and had countless alternatives on quiet streets, parkways, multi use paths.

  18. Matt says:

    Here in Minneapolis, there is a push by the city and the county for bicycle safty, with the push for bicycle boulevards and now cycle tracks. These measures began years ago when we had multiple bus/bicycle deaths, where the results are that now the busses have the right of way over all forms of transportation, and bike lanes are designed to put bicycles between the busses on the right and on coming traffic. Instead of teaching bus drivers that they can’t kill people. That’s a whole different subject. What, the bicycle boulevards and particularly the cycle tracks are designed to do is give motorists an out of site out of mind mantality. Which will make it even more dangerous for cyclist who have to use busier unmarked streets, which you’ll hear more of “Use the bike path”. Also, it was stated by a City of Minnepolis bicycle and pedestrain representative that all of the boulevards and cycle tracks are being designed for 60% of the cycling community, that 60% are your fair weather riders who would rather be on a bike path than in traffic. There’s about 7% of actual cyclist that ride all year, up here in the tundra, whom ride more miles and hours than the 60%, and most streets that get plowed fiirst are busy streets. I have a hard time riding through 15″ of snow that won’t get plowed for three days like on the boulevards and cycle tracks. I like the bike lanes on busier streets, but they can be over done. I see a backlash coming from all this caring.

  19. Mike says:

    2 comments:

    1. Bicycle boulevards work in a grid system, but the newer sections of almost all towns and suburban developments are NOT gridded. Take a look at the map for any city – the urban core is usually a grid but everything else looks like a nest of worms, with cul de sacs galore that feed into a few arterials. We need to change this design paradigm and go back to the grid.

    2. I take issue with your statement that “if the traffic speed is too high for a painted bike lane in the roadway, but there are too many intersections for a separate path, then the speed limit should be decreased to make the painted bike lane safe. Fortunately, traffic speeds have to decrease as the intersection density increases, so the needs of cyclists trend in the same direction as those of other traffic. Decreasing traffic speeds a little further will make things safer for all.” It has been shown that motorists will drive at what they percieve to be a safe speed for a particular road, no matter what the posted speed limit is. For this I refer you to a Michigan DOT paper, “Establishing Safe and Realistic Speed Limits. “The majority of drivers, consciously or unconsciously,
    consider the factors in the driving environment and travel at a speed that is safe and comfortable regardless of the posted speed limit.” See http://www.michigan.gov/documents/Establishing_Realistic_Speedlimits_85625_7.pdf

    Therefore, if as you state the speed limit is reduced to accomodate bike traffic, but to the driver it appears that the road is designed for a higher speed, this lower speed limit will be largely ignored. Better to separate the bikes from the cars.

    • I am all for separating the bikes from cars, but if there are too many intersections, you cannot do that. You can provide a semblance of separation, but you still have the mix of bikes and cars at the intersections.

      Speed limits can be designed into the roadway, rather than just posted on signs.

  20. Mike Beck says:

    This plan is an idealized musing of city planning, Jan.

    Bike lanes only on streets 30mph and below, separated cycle tracks for roads with any faster speeds, but the neighborhood greenways don’t have enough of a bicycle focus to provide effective or safe bicycling. My, you have a very exacting set of what makes traffic corridors suitable for bicycling – your way.

    And some of the bicycle planning seen and lauded in Seattle isn’t good enough, it’s going to merit even more separation……..Seattle’s put in bikelanes on previously 40mph roads, done road diets and given space on the road to bicyclists, but traffic still moves 35mph. On these roads, found across Seattle, your plan implies ripping out the road diet and bike lane, to add cycletracks. The city isn’t going to reduce the speed of traffic on 15th in lower Queen Anne, so this is a road that apparently merits a cycletrack stuffed in there somewhere, in your opinion. And, it really does. Somehow.

    Paradoxically, Jan, your vision of an American cities chockablock with 30mph streets with bikelanes, bike boulevards, and cycletracks where speeds are higher sounds – well, remarkably dutch of you.

    Look at Bellevue, Beaverton, San Diego, Kansas City, Austin, Boise, Cheyenne, Atlanta, Charlotte….wherever you look across the US, there’s a lot of roads with traffic speeds over 30mph, and they’re not going to be reduced to benefit the cyclists.

    Your plan, Jan, suggests an extensive system of separated bikeways for places like Bellevue or San Diego, for any suburb in America for that matter. All the political bicyclists can muster isn’t going to turn suburban arterial roads into 30mph traffic safe roads with bikelanes.

    I like your plan of traffic separation, mixed traffic and bikelanes all rolled into one city plan, though. Excellent idea. I wonder if Seattle’s thought of this yet?

    • The current draft of the Seattle Bike Master Plan already consider the mix of facilities that I (and others) propose. However, the Master Plan does not consider intersection density as a design criterium. And that is a recipe for more accidents. These blog posts have created a very positive discussion with several city council members. Hopefully, the draft will be changed based on this input, and also based on the experience with the first, flawed cycletracks that the city has installed.

      Since you mention Bellevue, there already are bike routes through Bellevue that come close to “Bicycle Boulevards,” with a few connectors that have bike lanes. One key component of any bicycle plan must be an easy way to find good routes. Nobody suggests that cyclists should mix it with high-speed traffic on busy corridors. Most experienced cyclists know routes that are safe and pleasant, but novice cyclists often don’t, so they ride on Aurora and other streets that may be reasonably safe, but certainly are unpleasant.

      I don’t think you’ve lived and cycled in a place with poorly designed cyclepaths. I have, and I don’t want to go back there, nor do I want those poorly designed facilities come here.

  21. Erik Griswold says:

    You are incorrect in saying that Copenhagen grew organically. Yes, the old city that you show on that map is medieval but everything outside the old walls has been built since 1869. So most of the Copenhagen you experience is actually younger than Manhattan.

    • You are right about the history. However, even the newer parts of European cities don’t use a grid layout like older American cities and towns. That said, they found room in Munich for “Fahrradstraßen” (bicycle boulevards), so perhaps the concept will be adopted elsewhere in Europe as well.

  22. Wayne Pein says:

    If there are multiple lanes, motorists should change lanes to pass, as they would for stopped busses or front loaders or for slower cars, etc. The rightmost lane is commonly called the “slow” lane. Why should bicycle users be subject to a 4′ “lane” that is barely wider than their essential operating space? If I drive my motorcycle, I get a full standard lane width as buffer from other vehicle drivers. Why should I be relegated to a second class road user just because I opt to BE the engine?

    Further, the argument that a bike lane legitimizes our right to the road is as bogus as saying a “Colored Bathroom” legitimizes black people’s right to urinate. The placement of a bike lane, no matter how poorly engineered, means one must use it either by law or by the “law of the jungle,” aka motorist coercion.

  23. Mike Beck says:

    Planning for bike traffic where people want to travel is a key component to moving people by bike. Your last comments about suburbia read like you’re willing to surrender the main arterial suburban roads to the motorists and leave cyclists to find their way on little traveled side streets.

    Very picayune, and quite repressive to bike travel.

    One of the problem of suggesting there’s nothing to be done on roads like those found across the typical, high speed, suburban sprawl distinctly unfriendly to bike traffic is that a lot of suburban america has no slow speed street network with which to carve out bicycle boulevards.

    You seem to suggest that the main roads in suburban cities like Bellevue should remain mostly as is, the domain of automobiles, and force cyclists to find their way on little used and out of the way streets. Sure, develop some bike boulevards, but no sense really, trying to fit bikes into the Bel-Red traffic corridor, for example. Too high of speeds, etc.

    If there’s a road with width AND speed, common across suburbia, a buffered/green lane modern US cycletrack has strong potential to help normalize biking along roads, vs shunting bike traffic to shaded side streets that don’t actually go where people want to travel.

    Again, i’m not sure your traffic musings present solid solutions to mixing bike and automobile traffic in cities across America. Esp at that magic ‘over 30mph’ mark, where you call bikelanes unsuitable, and need more significant separation from auto traffic.

    Calling for robust traffic separation at 30mph and above, wether you believe it or not, is a call for a lot more cycletracks and green bikelanes in America, or bikes shunted off the faster traffic corridors altogether.

    Jan’s typology of bike planning is a far stronger call to clear the roads for cars than planning for cycletracks/green lanes.

    Jan, It honestly makes me wonder how you ride Bel Red traffic corridor, if you want to go from Bel Square to the Performance bike store on the Eastside –

    For those of you riders unfamiliar with Seattle’s suburban east side neighbor, Bellevue, look it up on google street view. There’s NO slow speed streets to turn into bike boulevards to get between the two- all the roads are high speed arterial roads. I have a sneaking suspicion a lot of the US may be ‘off limits’.

    To execute Jan’s plan for robust traffic separation at speeds over 30mph, most of Bellevue’s roads and the majority of suburban america call for -you guessed it – cycletracks!

    I find it interesting how Jan’s position on facilitating bike traffic, in its eager yet inchoate way, shrouds an ill-defined but nonetheless crystalline call for the build up of robust cycle track networks across suburban america.

    • I totally agree with you that on many suburban roads, cycletracks may be a good solution. High traffic speeds, low intersection density.

      Right now, I see cycletracks implemented mostly in dense urban neighborhoods, with high intersection density. They often put cyclists on the wrong side of the street. I don’t think that is a good idea, and that is the reason for these posts.

      If you are in favor of cycletracks, it’s in your interest to make sure they are implemented well and safely, so they don’t get a bad name before they even have a chance to spread to the suburbs.

  24. Zbyszek Kolendo says:

    Oh Mike, please! Making Jan answer the same arguments over and over again may keep him too busy to come up with some other brilliant observations he’s so good at. And we do not want that, do we? -:)

    Zbyszek Kolendo
    Gd, Pl

  25. Mike Beck says:

    Cities can restrict turns across cycletracks to reduce conflicts. This has been done in Vancouver BC, NYC and San Francisco to great effect on their urban cycletrack/green lanes.

  26. Opus the Poet says:

    I keep reading articles like this from where I am in TX, and it makes me jealous. In the state of TX the lowest speed limit that can be legally posted on a public street or road is 30 MPH, meaning every residential street should be equipped with cycletracks by the data you presented. Seriously our residential streets are 30 and 35 MPH, with arterials 40-50 MPH with the residential arterials on the low end of this range but 45 MPH residential arterials not unknown or even uncommon. These are speed limits, but the actual speeds are generally 10-15 MPH higher, so we get freeway speeds through residential areas. Well what would be freeway speed in most other states, here we just raised the freeway speeds to 70-80 MPH. Combine that with zero accommodation for any mode of travel other than motor vehicle and riding a bike is a endless series of white-knuckle moments.

    • In the state of TX the lowest speed limit that can be legally posted on a public street or road is 30 MPH, meaning every residential street should be equipped with cycletracks by the data you presented.

      You always need to start with intersection density. If the speed limit cannot be reduced, you can design the streets to reduce actual vehicle speeds. When I lived in Austin, TX, that was done by putting a stop sign at every intersection. Crude, wastes a lot of gas, but effective.

      Of course, in many places, you’ll have to compromise if there is no political will to design proper facilities (and proper speed limits). However, if you compromise, a poorly designed cyclepath is the worst option for cyclists. It gets bikes off the road and out of the minds of drivers, until the next intersection…

Comments are closed.