Wednesday 12 June 2013

Ontario Traffic Manual "Bicycle Facilities" draft edition. How not to design for cycling

Someone just sent me a link to the "final draft" of Ontario's new Bicycle Facilities manual. This is book 18 of the Ontario Traffic Manual, covering all aspects of design for all modes of transport. I have not read the other manuals and can only hope they are better than this one because unfortunately from the front cover page onward the manual for bicycle facilities demonstrates a remarkably low aspiration for cycling.

The good bit
Early in the manual people are divided into four potential types of cyclist and the authors recognize that 60% are "Interested but Concerned". This is a reflection of a lack of subjective safety, a theme covered many times on this blog.

It is a good thing that Ontario has identified that future cyclists can only come from that part of the population who do not currently cycle. In order for cycling to grow, the experience of cycling has to become more acceptable to the masses who do not currently cycle. This is why subjective safety is important. Cycling has to not only be safe but it must also feel safe. For this to happen requires the frequency of conflicts between cyclists and motor vehicles to be reduced. It requires this to happen no matter where the cyclist is riding from or to.

It is also necessary for cyclists' journeys to be efficient. It is often possible for cyclists to make shorter journeys than would be necessary by car or to have to stop less frequently for traffic lights than they would by car. It is possible for cyclists' journeys to take direct routes which are not in the same places as those taken by motor vehicle. It is possible for cycling journeys to be pleasant free of stress. That's the reality that we live in and it's what this blog tries to present to the world.

But does Ontario actually intend to take this on, or are their proposed interventions too minor to make a difference ?

What the document actually says...
The language of the document is slippery. Some of it sounds quite reasonable on an initial reading, but when you look closely it becomes obvious that the authors have rather low aspirations for cycling. There is an expectation that cyclists can share the roadway when both speeds and traffic volumes are at higher levels than we would experience. The authors think that it is only necessary to "consider" building an on-road cycle lane even with speeds of up to 100 km/h. The language of the document betrays the lack of ambition for cycling.

Diagrams like this have been produced across the world. However, to encourage cycling the bands need to be adjusted so that cyclists have their own infrastructure at far lower speeds and smaller traffic volumes than suggested here.
A passage which considers the needs of "Experienced cyclists (commuter or other utilitarian)" reads: "This group generally prefers direct, continuous facilities with minimal delay as is generally provided by the arterial road Experienced cyclists (commuter or network. Experienced cyclists may be comfortable on shared use other utilitarian) roadways with low motor vehicle volumes and speeds. However, users in this group typically prefer on-street bike lanes or separated facilities where the context warrants it." This sounds just fine until you consider what they're talking about. Our 85 year old neighbour's 80 years of experience is surely enough and her regular trips to the shops by bike are definitely utilitarian but I have never heard her express a preference for riding on arterial roads rather than the more direct, safer and more pleasant cycle-paths that we all use. Riding on road is equated falsely within this manual with making journeys which are direct and continuous and have minimal delay. There is no need whatsoever for that to be the case. There is a reason why fast cyclists in the Netherlands ride on the cycle-paths.

Similarly, when writing about "child cyclists" the authors say "This group generally requires separated facilities free of conflicts with motor vehicle traffic. Separated facilities should be considered near schools, parks and neighbourhoods. Children under the age of 11 should be permitted to cycle on sidewalks since they may not have the cognitive ability or experience to ride on roads with motor vehicles by themselves." This also sounds almost reasonable until you realise how restricting this is. The need that childrens have for separated facilities is to be considered only near these specific locations. The freedom of children is thus to be restricted to routes between places that the road designers have decided in advance that they might like to go to. Canadian children "under the age of 11" supposedly need to ride on sidewalks because they lack some unknown cognitive ability while Dutch children regularly make school trips by bike from a far younger age. Indeed, my daughter rode 150 km out into the countryside on a school cycle-camping trip at the age of 11. Nothing special - almost all Dutch primary schools do this.

There is actually no need for this compartmentalism. Children benefit most from exactly the same infrastructure as fast and experienced cyclists benefit from most. i.e. that which allows direct and convenient journeys to be made in safety.

These two passages demonstrate the lack of ambition of the plans for Ontario. But this was just scraping the surface. Just look at some of their recommended designs for infrastructure...

Is this bike really in a safe
place relative to the car ?
Something I've been meaning to write about for a long time is the strange North American fixation with painting "Shared lane markings" on streets.

This is not infrastructure, it's just paint. What's more, it's not even paint which attempts to give cyclists their own space. Sharrows do not place any physical object in the way of drivers. They don't even suggest that drivers should steer clear of cyclists. Sharrows are merely a tokenistic attempt to pacify cyclists. They are not a feature of roads in countries which have a high cycling modal share. Sharrows give the appearance that something has been done but without any real change having been made to the road environment.

In Ontario, as shown in this picture, cyclists are to be encouraged to ride no further than 1 m from the kerb and drivers are actually to be encouraged to overtake cyclists within the same lane. The manual explains this in black and white: "if the travel lane width is 4.0 metres or greater, passing may be possible". And what type of road would this be ? "roads with higher traffic volumes, low to moderate speeds (40 to 60 km/h) and frequent intersections or driveways."

Inferior to on-road cycle-lanes but sharing many of the same problems, sharrows are not the route to mass cycling.

Faith in signage

It's neither safe nor pleasant for a cyclist to be passed within the lane by a vehicle travelling at a considerably higher speed. So how does Ontario propose to deal with this problem ?

Please pass cyclists
within the lane...
Answer: erecting signs which read "share the road" alongside the sharrows.

This does nothing to improve the subjective or actual safety of cyclists.

For narrower roads with less than 4 m lane widths, Ontarian planners suggest that the sharrows should be further out and cyclists be used as mobile traffic calming devices behind which motorists will have to wait until they can pass in safety. This on roads with 50 km/h speed limits. i.e. a speed which only a small fraction of the population can maintain for any period of time (the world one hour record for a reasonably normal bike is still under 50 km).

If you have the guts to want to do it, then perhaps it is marginally safer to ride in front of frustrated drivers than to encourage them to pass within the lane. However it is unlikely to be a pleasant experience.

Drivers who are in a hurry, stuck behind a bike and tooting their horns haven't necessarily seen a sharrow and don't necessarily understand why you are obstructing them. In any place where cyclists are used as traffic calming devices the dream of mass cycling will remain a dream. This is not the hassle free cycling that is required to encourage the entire population to ride bikes. This experience will do nothing to convert the 60% of the population which Ontario has identified as being "Interested but Concerned" into regular cyclists.

Paved shoulders
The document refers to "paved shoulders". The recommendations for widths are vague: "should typically have shoulders between 1.5 and 2.0 metres of pavement width depending on the volume, speed and mix of vehicular traffic" is followed in the same paragraph by "practitioners may consider providing a minimum paved shoulder width of 1.2 metres after applying good engineering judgement and consideration of the context specific conditions." The lower minimum is available to anyone who thinks they have a "constrained corridor", but of course as we all know already, every place in the world claims to have "not enough space" and that includes Canada.
Sometimes they make it too easy to criticise. Of the three examples that Ontario provides of real paved shoulders, two have a car parked in them.
I'm happy to say that we don't have "paved shoulders" in the Netherlands. Or at least we don't have them as cycling facilities. The idea that cycling is made attractive by providing nothing more than a stripe of asphalt at the side of a road which may carry large volumes of high speed traffic is quite remarkable. It takes more than this. The optional separation by buffer of width 0.5 to 1.0 m is inadequate to lead to a high degree of subjective safety.

On Road Cycle lanes
We move on now to on-road cycle-lanes. These do exist in the Netherlands but they're generally older facilities and are not nearly so common as properly segregated cycle-paths (just 5500 km exist vs. 37000 km of segregated paths). You would expect to see such lanes in the Netherlands mostly on streets with fewer vehicles or occasionally in places which simply haven't yet been updated to remove them. They are found infrequently on busy roads or with frequently used parking because they are not suitable provision in such areas.

In Ontario such lanes are to be "1.8 metres wide, measured to the face of the curb or, in its absence, the edge of the roadway. Practitioners may provide a 2.0 metre facility on roadways with higher bicycle volumes to facilitate overtaking within the bicycle lane." and "bike lanes are typically no wider than
this so that they are not misinterpreted as being for general traffic use."

However, the diagram later on shows a slightly different idea. The desired width for a bicycle lane next to parked cars somehow shrinks to just 1.5 m, with a 1 m buffer (which may be reduced to 0.5 m) to protect from "dooring". This is simply inadequate. There also appears here a mention of the oft-mooted idea of cycle-lanes between lanes of cars. These apparently need to be just 1.8 m minimum in width.

However, on the subject of cycle-lanes, the drawings of proposed road layouts provide the most entertainment. Almost every one of their examples is flawed from the point of view of cyclist safety and most of these flaws are obvious at the first glance:
Problems caused by this arrangement include:
1. dooring due to opening of doors of parked cars
2. cars passing being too close.
3. cars wishing to park swerving across the cycle-lane
4. cars turning the corner swerving across the cycle-lane
5. Nothing is done to protect cyclists at the junctions - the most dangerous places for riding.
6. How does a cyclist make a safe left turn at these junctions ?
Added red and blue lines show how the paths of cyclists and drivers clash at a junction designed like this. It's just not good enough for lanes to disappear and signage isn't enough to prevent the problem.
In this example a cycle-lane disappears right at the most dangerous point - where the roadway narrows. The Ontarian designers place all emphasis on avoiding a collision on a cyclist who can turn his head by 180 degrees and judge the safe moment to pull left into other traffic. This is simply not a safe arrangement. I've added a blue triangle to show how kerbs on the road could be used to make this merging safer by forcing drivers to pull to the left.
Another example of designing in conflict. Cyclists heading straight on should not expect drivers turning right to merge into their lane. This arrangement of lanes requires both cyclists and drivers to be able to see behind themselves to avoid colliding with one another while they change lane while they must also look forwards to see what vehicles in front of them are doing as they may be adjusting their speed to merge or slowing for a red traffic light.

From the point of view of sustainable safety  this design is extraordinarily bad because it relies on everyone behaving perfectly at all times. For this to work without injuries, no-one must ever be tired or distracted. It's the precise opposite of the Dutch principles which seek to make roads safer despite their users' misbehaviour or mistakes and self-explanatory so that users do the right thing without having to take notice of excessive signage.
These examples were all found within the first third of the manual. There are far more examples of bad design which I have skimmed over and while it would be amusing to go through all of them this would also be very time consuming so I'm stopping here (unless Ontario wants to sponsor me to continue, that is).

Conclusion - please think again
I've read a few design manuals in my time and unfortunately I have to say that this is one of the very worst. The examples above come just a third of the document. Flipping through more of the pages reveals seemingly an endless stream of material which is just as bad and in some cases obviously worse than that above.

It appears that no stone was left un-turned in seeking out bad ideas. Amongst these further bad ideas are two stage turns, bike boxes, cycle-lanes in the middle of the road, "jug handles" to make left turns, recommendations for separated paths to be too narrow, bus stops on cycle-paths, paths shared with pedestrians, awful ideas about what to do with cyclists at roundabouts, junction design which very probably will be proven to be "to die for" and the suggestion of signs, signs and more signs to try to explain the whole mess to the users of the infrastructure

Take advice from people with experience. Learn from best practice.
Ontario: don't accept this document as your future guide to building bicycle facilities. It is inadequate to the task in more ways than I have time to document.

If your aim is to achieve a higher cycling modal share, encourage a wider range of the population to cycle and to improve the safety of your cyclists then you desperately need to start again on your manual.

Instead of seeking out only inexpensive "solutions" or inventing new ideas, and instead of being influenced by countries which have similarly low cycling modal shares to your own, please take a look first at what has been achieved here in the Netherlands. It makes absolutely no sense at all to ignore the most successful practices.

A good start would be to buy copies of the CROW manuals and read them thoroughly. These are the best documents that you'll find anywhere about cycling infrastructure. However, don't stop at just reading books and websites and be aware that the real life experience is only hinted at by what you'll read. You also need to know how it feels to use the best cycling infrastructure in the world and you'll find out how the best infrastructure explains itself and does not need the number of signs that you are planning to install.

You need to aim high in order to achieve future success. This is the reason why we offer cycling infrastructure study tours - so that professionals such as yourselves can see what best practice is and avoid making expensive mistakes.

A pattern...
A few days after writing this, I realised that the same consulting company was involved as had been paid to put together another inadequate plan. Not every organisation which claims expertise actually has it.

I really am willing to go through the rest of the manual for a fee. It could be instructive to point out all the problems. However, I think time would be better spent by Ontario setting about writing a proper manual for cycling facilities.


chrismealy said...

Seattle is having public meetings on its next bike master plan this month. Here's a recent draft. I would be incredibility grateful if you did a post on it.

somebody said...

Thank you for a thoughtful and clear review of this manual. It appears to be very similar to the bike infrastructure being built in the San Francisco area and probably a good representation of North American cycling planning at this time.

I completely agree that the "infrastructure" being built is generally just some paint on roads (and usually the most heavy car traffic roads!) and won't enable many more people to cycle. I would describe cycling conditions in North America as an extreme sport and it appeals only to the very brave and confidant cyclist.

I think the fascination with the "sharrow" in North America is that it is used as a driver education device that a cyclist is actually legally allowed to be in the road. They seem to reduce the amount of road rage and honking. I don't think it does much to improve collision safety and probably has some negative safety aspects as well. Unfortunately it seems the "share the road" sign is often literally taken as occupying the same lane at the same time!

I hope that North American cycling facilities evolve quickly to separate bike and car routes. In the US we have a bit of a history with the word "segregation" so I would suggest a more neutral word such as partition.

David Hembrow said...

Chris, I took a quick look at Seattle's plans and there are certainly things that should be criticised. Unfortunately, doing so would be very time consuming so I don't promise anything soon. I did do a post about Los Angeles a little while ago.

Reclusivity: Thanks for your comment also. I'd read of problems of people interpreting "share the road" as meaning the same lane, but Ontario's official sanctioning of the same was a shock to me.

I know that Americans have a problem with the word segregation and I understand exactly why - but this is problematic as it's the correct word to use.

What's more, to use the word "partition" instead could also be problematic. I'm afraid we all have to face our history.

Clark Nikolai said...

Not to mention the attitude that some hold in North America where the disenfranchised are framed as privileged. Any crumbs thrown to them is seen as unworthy perks.

This comic is a parody but it shows the sentiment.,26723/

Bob said...

Well David, as someone who now must live in the province of Ontario where I see this silliness first hand, all I can say is, you've done a very fine job of expounding on a notion of bicycle traffic management that I would have simply summed up by using the term "Bullsh*t". There have been a couple segregated bike lanes opened here and there (to much fan fare I might add) but the bottom line is, all they do is paint some lines on the road, and figure *that* should take care of it.
Traffic is at a stand still every single day, and nobody in 30 years has had the political fortitude to deal with our failing traffic infrastructure. Bike traffic seems to get nothing more than a few lines on the road. Sad, really.
Mostly just a waste of paint, since it does damned little for the safety of bike riders.
Ontario is our home, and we moved back to it, but if we didn't have a house/job here, I'm not so sure it would have been our first choice.

Jim Moore said...

Given your connection with NZ you might be interested, if you haven't seen it already, in the "Christchurch Cycle Design Guidelines" available at

The background for this document is of course the two recent tragic life-taking earthquakes. The City Council and the citizens, at the same time as having to rebuild their city, has also given due consideration as how people and goods will be moved with a particular focus on cycling.

Their 2013 cycling guidelines are a mix of Dutch (intersections), Danish (mid-block) and the Anglosphere (bike boxes, shared paths and quiet routes for inexperienced cyclists).

bz2 said...

What's wrong with a jug handle left turn? If spaced widely enough, it lets the cyclist and motorist approach each other at right angles, which is probably the most important factor in safe junction design...

Unknown said...

Specifically which CROW manuals should we purchase? (for cycling advocates abroad?)

David Hembrow said...

Clark & Bob: I hope things get better for you in Canada. There's really no reason why they should not. Political will is largely a function of what the public asks for and so of course public opinion is something that needs some work. I think it's important to emphasize that bike infrastructure is not just for "cyclists", but also for other groups that might be more important such as those who find it more difficult to get about, older people, little people. It's even good for those who drive cars.

Jim: Thanks for sending the Christchurch guidelines. Looks like a bit of a mixed bag as always, including cycle-lanes between lanes of cars, advanced stop lines, two stage right turns and several other rather dodgy ideas. Basically it looks like they've drawn versions of everything that's been proposed or built everywhere.

I see they've even picked up on the idea of what they call a "cycle Barnes dance" from my blog. I can tell because they nicked one of my photos from here to illustrate it, editing off the text in the corner which says where it comes from. I suppose that's a positive sign of sorts, but their "design concept" doesn't really look like a Dutch junction.

All too many planning "experts" these days do their job by scooping up photos off the internet and pretending they understand them isn't really good enough. Our tours are offered specifically to help people like this to avoid making expensive mistakes.

bz2: The problem with a jug handle on a road with two lanes of traffic alongside it is that instead of a cyclist being able to make a simple turn across one lane without stopping they are asked to stop, turn their bike and cross two lanes at the same time. It can work in some circumstances, but not as shown in the Ontario manual.

David Hembrow said...

Andrew: The Design Manual for Bicycle Traffic and the Road Safety Manual would be a good start. You can find both here.

But be wary of anyone who has only read manuals. The reality on the ground here is a little different to what you'd be likely to build if your only experience was these manuals.

Anonymous said...

One thing I am curious to learn, is how did things begin in the Netherlands. The infrastructure we have in Ontario is pathetic. These cycling plans are poor and often dangerous. Only the most dedicated are willing to risk riding most routes. Here in Guelph, my wife has a route that is almost safe. A big bike lane takes her 90% of her trip. Perfect except for this one section where her nice big bike lane disappears into a very tight narrow two lane road with loose gravel shoulders and ripped up pavement. Now she has several different travel paths to choose from, and everyone but one has this same design flaw, bike lanes that channel her into the worst possible road section to have to share with cars. The only one without this gap in “safe” space is to take the busiest most traffic congested route. My routes have no bike lanes. And again the “safest” route has the most, and fastest traffic, but also the most lane room.

As for taking the lane, as is advised, when I was forced to do so on occasion because of the unbelievable aggressiveness of drivers I am sharing a lane with, I’ve been met with drivers pulling into the second lane not to safely pass me, but to then slow down and yell at me for taking up a lane for my own safety, asking “what is my problem?” One was even thoughtful enough to try and “teach me”, whipping in front of me to slam on his brakes. A sadly common aggressive driver technique to “teach” other drivers not to get in the way by dangerously passing them only to then force them to slam on the brakes to prevent rear ending the “teacher”. A very effective and dangerous technique when used on another driver, much less effective against someone doing 25kph in a 60 zone.

But as poor as the sharrows, lane sharing, and bike lanes are. They are a beginning. A poor and slow start to creating proper cycling infrastructure. But it is a start, and until sufficient public support is in place to create the political will to change the culture from the car-centric focus currently in place, nothing more will occur. If a linked network of bike lanes can create an increase in adult commuting cyclists, and sharrows can help build subconscious acceptance that bikes will be on the road, then that shift can begin.

Here in Guelph, there is praise for the cycling lanes we have. I don’t understand the praise as the lanes are fractured and consist of lanes as short as 15 meters that go nowhere from nowhere. Still these lanes are a beginning, and if in the following years, these fractured paths become an interconnected network then at least in this town there will be the infrastructure in place to build a larger cycling community to have the political power to begin building better infrastructure.

Thank you for reviewing the Ontario Traffic Manual, can’t say I expect it to have an impact on the powers that be, but it should provide help to those community groups trying to change that staus quo.

Nick said...

Thanks for the post. Here’s the problem as I see it:

If the Ontario Traffic Manual for Bicycle Facilities featured the best practice of Dutch design and set high standards for when to apply such facilities, it would be REJECTED by the Province.

Design guidelines like the ones you criticize here are designed to help get things made. There are a million little reasons why standard dutch practice are incompatible with the other traffic manuals in play in Ontario. (They aren't good reasons, but they are real reasons).

Clearly, the whole approach to transportation in Ontario needs an overhaul, but I doubt the province takes bicycling seriously enough to have this traffic manual be the start of that process.

Please continue to beat the drum and critique work like this on it’s merits as you do. As a tool for designing the best bikeways in the world, this document absolutely falls flat. But as a tool to help move Ontario forward, if only a little bit? It might be just what the city needs.

Anonymous said...

We once did a tour in New Zealand where we learn from the aggressivity of car/truck drivers that cyclists are supposed to drive on the lane shoulders. We got horned and insulted until finally we got the point.
But we did not drive on the shoulders even after knowing it. Why?
Lane shoulders in New Zealand were, we guess, designed and built for mountain bikes with full suspension, not road/trekking bikes. Second, the shoulders have been used as waste disposal (especially empty beer bottle glass) by car/truck drivers. So the shoulders were littered with broken glass! And finally, sometimes the shoulders were barely existing, either because bushes invade them or because the road designers thought that 5cm was enough!
My conclusion from that experience is that lane shoulders should be a no-go for cycling zones!

Note: Apart from these little inconveniences, we had a great time cycling in New Zealand.

Anonymous said...

David, have you already seen the "Cycle Infrastructure" book ( What do you think of it?

David Hembrow said...

Alex, no I've not seen that book. If it shows highlights then it's missed the point IMO.

Anonymous said...

Yea, I got it.
I'm thinking of translating some cyclepath design manual into Russian. Most people involved in road design and constructing in Russia unfortunately don't read in English. What make it even worse, they aren't interested in such kind of books. If there was a demand from professionals I would obviously choose CROW's "Design manual for bicycle traffic".
Now, however, I think, we need information source aimed rather for wider audience to form common view and request for good cycling infrastructure.
I have in mind few possible options:
- NACTO Urban bikeway design guide,
- PRESTO guides and fact sheets,
- Dutch "Collection of cycle concepts".
I would be very grateful if you may give me a hint on some data sets either from this list or some others, that may be helpful for professionals and understandable for whole cyclist community.

David Hembrow said...

Alex, please forget about NACTO. It's a hopeless guide to bad design ideas and will keep you from achieving what is possible.

Not seen PRESTO.

CROW's guides are as good as they get, but note that they don't give any context. They're very passive in language and I think can be misleading to people who have not seen how the guidelines are applied in reality.

For example, CROW is very passive about what is good and popular and what is less good and mostly not built.

This is why we suggest that they should be read and absorbed, but that people who read them also benefit greatly from coming on one of our study tours so that they can see how the infrastructure really is built and really works.

Anonymous said...

Thanks, Dave.
So, NACTO goes to a trash bin. Choice is easier now :)
Here is PRESTO —
I know about your cycling tours. The idea is great. But (I know it's typical for Russians to start a sentence with "but") I don't know anyone from the city administration or road planners to whom it may be worth to suggest to take a part in such tour. All of them are anticyclists.

David Hembrow said...

Alex, I took a quick look at PRESTO and realised I have seen it before.

I don't have time to go into details, but I'd say it's a mixed bag. Some parts are perfectly fine, others are overly compromised. It is perhaps not surprising, given that it's a document written by a large international committee, that the recommendations have had to take into account people who don't always have the best solutions.

PRESTO is not completely bad, but CROW is still your best bet IMO.

I still think CROW is not enough on its own, though. An "anticyclist" could design truly bad infrastucture with nothing but the CROW manuals as his reference. It simply requires that the reader picks all the things from CROW which look like they're least expensive and takes all the recommendations as maximums. The result would be something which had apparently come from the CROW manual, but didn't really resemble Dutch infrastructure very closely at all.

You must be very careful about "Going Dutch" because many things which don't really resemble Dutch practice at all are now sold as "Dutch".

Multiparty Democracy Today said...

I don't know what the Dutch would call a cycle path in the non built up areas where pedestrians really do walk on the cycle paths. That is one of three places I would ever give pedestrians and cyclists mixed space. The second is a pedestrianized street. The last is in a woonerf without a sidewalk. The widths are not good either. 3 metre paths are secondary routes and shortcuts, not main routes. 1.8 metre wide cycle track is pretty bad for a one way. 2 metres is the minimum (emphasis on minimum) and 2.5 is the standard.