Monday 2 April 2012

100% segregation of bikes and cars

City centre street, no cars allowed. Clear signage gives
loading times for deliveries
Again and again I read comments of the form "even in Holland not all parts of rides are segregated" or talk of "similar road exposure" in the UK vs. the Netherlands. I've been accused of making out that the cycle-path network in the Netherlands is more extensive than it actually is.

Such comments are based on a misunderstanding of the differences between the way that roads and streets are designed and used in the two countries.

"Shared" street ? Not on an equal basis. It's a through route
for bikes but access only for cars.
The misunderstanding usually arises because someone notices the simple statistic which says that while the Netherlands has about 130000 km of roads, there are only about 35000 km of cycle-path.

This sounds, of course, like there is a serious short-fall of segregated cycle paths.

The first misconception is that people look at these figures as if those 130000 km of roads are exactly the same as roads in their own country and assume that cyclists find themselves in conditions which are like in their own country. That's not correct.

The second misconception is to assume that because there are a quarter as many km of cycle-paths as roads that Dutch cyclists spend 3/4 of their time riding on roads. That's also not correct.

City centre streets are busy, but with bikes instead of cars
Quite simply, these are the wrong ways to look at those figures.

Dutch people do cycle on a mixture of cycle-paths and roads. However, this does not mean that cyclists in the Netherlands spend much of their time mixing with motor vehicles.

Over the last few decades, the Netherlands has unravelled the networks of car and bicycle routes. If you compare routes for the same journey by bicycle and by car, then in very many cases you will find that the two routes are very different to one another.

A child rides a bike in the middle of a bicycle street.
This was once a busy through road by car. Now access
only by car / through route by bike.
I've demonstrated this principle several times on this blog (1 2 3 4 5 6 7 8). In some cases the separation of routes is achieved by building of cycle-paths. However, in many others it is achieved by removing cars from the roads. Sometimes when new roads are built, the old road can serve cyclists very well.

These days, very many roads and streets are simply no longer part of the route network for cars. They still allow access by car, so that deliveries can be made and people who live along them may reach their own homes by car should they wish, but they are not through routes by car. As a result, such streets are absolutely not dominated by cars. A cyclist using a road on such a route has much the same feelings about safety as someone using a cycle-path.

It is only when cycling doesn't feel like an extreme sport that it can become so popular as it is in the Netherlands.

Residential street. No cycle-path required as it's not a
through road for cars. Residential parking, but these cars
rarely move.
Residential streets in the Netherlands rarely work as through roads for cars, even if they were originally designed to do so. This makes them excellent places to cycle or walk with a high degree of comfort and safety.

In some cases, this has been formalized by creation of woonerven, but many of the same characteristics can be found in residential streets which are not woonerven.

Woonerven have a speed limit of "walking pace". However, they are not the only roads with low speed limits. In fact, over 40000 km of roads in the Netherlands, a third of the total, have a speed limit of 30 km/h or lower. This lowering of speeds on minor roads has been done to the maximum possible extent. It is now difficult to achieve more safety by this method  as roads which remain through routes for motor vehicles are not effectively calmed simply by changing the speed limit.

In a village, primary school children cycle home
unaccompanied on streets which are not through routes
for cars. From an average age of 8.6 children
travel unaccompanied
The photos give examples of different roads and streets on which cyclists have the same feeling of subjective safety, and similar degree of actual safety, as on a cycle-path.

These principles are not used only in towns. They are also common in villages and in the countryside.

Minor streets in villages don't work as through roads so drivers don't use those to "rat run" either. This makes it possible for quite young children to cycle to school and back unaccompanied.
Between towns in the Netherlands you can often find two roads next to each other. One for cars, the other for cyclists, agricultural vehicles and access to homes. Sometimes the route for drivers isn't visible from the route for cyclists.
Route signs in a village. Only one direction offered by
car, lots by bike.
Even in the open countryside you can ride long distances by bike on "roads" and rarely see any cars at all. Such roads only make a through network by bicycle because they are joined by cycle-paths. For drivers, many roads in the countryside offer nothing more than long detours, dead-ends, and roads with rough surfaces which don't get cleared in the winter.

Unravelling, or separating, the routes taken by motorists and cyclists has many advantages, including a reduction of noise and exhaust fumes as well as the increase in safety which comes due to a reduction in the number of chances for conflict.

Cyclists achieve subjective safety when they do not have to mix with cars, not specifically because they are on a cycle-path. Roads which have (almost) no cars on them do not require a cycle-path to parallel that road. A remarkably large proportion of roads in the Netherlands are, for all practical purposes, free of cars. This is just one reason 35000 km of cycle-path does not represent a short-fall compared with 130000 km of road.

Those roads on which there is an appreciable volume of traffic can be relied upon always to have segregated cycle-paths alongside. By this manner, very nearly full segregation of modes is possible when cycle-paths have a total length which is "only" about a quarter of the total length of the roads.

While cyclists are segregated from traffic in every photo in this blog post, this is the only photo in which this was achieved by building a cycle-path. This is a through road for cars and is used by a higher number of motor vehicles, so it needs a cycle-path.
The title of this post is deliberately provocative. However, it's not inaccurate. This has been my experience: When cycling in the Netherlands, you are almost always segregated from traffic, whether by means of a cycle-path, or otherwise. I have ridden many tens of thousands of kilometres in the Netherlands but not yet found anywhere that felt so dangerous for cycling as did my daily commute in Cambridge. My total experience of "road rage" in this country remains just one minor incident in nearly five years.

This map, which accompanied the printed version of this article from the Fietsberaad about Enschede, shows how main routes for cycling have in large part been unravelled from main routes for motor vehicles in that city. In the article it is explained that the absolute maximum number of motor vehicles per day on these routes should 2500 - i.e. a number that you might find on a residential street. More than this and the route starts to feel unpleasant and threatening for cyclists.
This unravelling of the cycle network from the motor vehicle network is taking place all across the Netherlands. Where there are higher levels of motor vehicle usage, even within residential areas, the public shows a preference for separated cycle-paths.

Many cyclists who visit the Netherlands on holiday remain oblivious to concepts like this. It is common that people who visit on holiday report that they had few problems cycling in the Netherlands despite there being cycle-paths for only part of their journey. Often this is ascribed to better driver behaviour, perhaps through better training. However, people who make such statements have simply not noticed this "hidden" policy. Roads here are not the same as roads in other countries. Segregation of modes takes place even where there are no cycle-paths. Sustainable safety principles require that conflict is reduced and that is what results in the much better subjective safety when cycling.

Our unique experience of having lived, cycled and campaigned in both the UK and the Netherlands has led to the programme of our three day study tours in which we transfer as much as possible of this knowledge. The tours are hands-on. Or, rather, they really are tours. You use the facilities for yourself, accompanied so that we can explain about what you are experiencing. To find out more, please book a tour.

It's not a substitute for a tour, but to get a flavour of what this is like in practice, please watch the following video. It shows a complete, normal journey by bike in the Netherlands. In this case, nothing more than the route from a fairly close by shopping centre to our home:

This post originally referred to 29000 km of cycle-path in the Netherlands. However, this was an older figure. There are now reckoned to be about 35000 km of cycle-paths.

This blog post introduced the term "unravelling" into the English language as a translation of the Dutch term "ontvlecthten". There was no previous English translation of this term. Others have picked up my translation while an academic paper in 2013 used the term "unbundling" instead to represent the same concept. 


Bob said...

And not only that....
but I'd like to add, as my neighbour (back in the Netherlands, in Delden) pointed out one day, if you're driving a car, and "A bike falls out of the sky", it's YOUR fault.
I thought those were good words to live by.
I also thought that was quite brilliant.
Bike riders and pedestrians are particularly vulnerable as it is, so when there is no separation, there should be no allowances for foolishness at all.
I hate to say it, but I don't hold out a lot of hope for Britain, or any of the other countries where the four wheeled beast is predominant. I have to include my native Canada in with that sorry lot.
It just seems like such a hopeless case, in spite of all the "Boris bikes" and their ilk.
And painting lines on the roadway? Please.

David Hembrow said...

Thanks for your comment Bob, but I must point out that it isn't quite true about the bikes falling out of the sky thing.

Bob said...

Well that (article) certainly makes sense, and I do seem to recall reading that very entry right here. I am getting somewhat forgetful I'm afraid.
I think perhaps my neighbour was being overly dramatic for the benefit of the "foreigner". That would be me.
Just the same, it was something that I certainly took under advisement when at the wheel of our car, even if he was pulling my leg ever so slightly.

H.Edink said...

Not everything is sunshine in the Netherlands. Last year I found a bad road for cyclists:,5.63538&spn=0.003103,0.005059

You may drive on this road 80km/h.
The sign warns for bikes on the road.

I admit that I'm a dutchman and this road is the worst ever seen in my country.

David Hembrow said...

H. Edink: As it happens, I've cycled there too. Last year I cycled between Flevonice and Lelystad Airport.

It's still the only example I've seen of somewhere which needs a cycle-path and doesn't yet have one. It is, of course, in what is so far the least densely populated part of an area of land which is part of the world's biggest artificial island. This area is not yet fully developed.

And while this is as bad as it gets, elsewhere I've ridden on far worse roads, with far higher volumes of traffic at higher speeds than you'd ever meet here, and no plans at all to do anything about it.

Anonymous said...

That is quite a a strange one, indeed there's even a cyclist here, almost on the verge:
I wouldn't fancy being overtaken within the lines there!

I'm not sure it's the absolute worst example in the country though. There seem to be a lot of roads in Limburg which have either no cycle provision or narrow lanes, which elsewhere would have segregated paths.

BC said...

In case you haven't seen this:

Sharing or Separation: Which Way for Streets of the Future? A debate with Ben Hamilton-Baillie and Steve Melia
Tuesday 22 May 2012
17:30 - 19:30

Jim Moore said...

Thank you David for this great blog piece. I'm just wondering, do people who come on your tours from low cycling rate places comment on the quietness of The Netherlands? To an Australian, watching the video at the end was quite eerie.

I would love to come on one of your tours - maybe next year with a bit of luck. All the best.

David Hembrow said...

BC, yes, "Shared Space" is something I specifically didn't mention in this piece. This is something which was tried with some enthusiasm in the Netherlands a few years back, but it was not a success in busy areas and you won't find new examples.

Ten years behind the times, Britain is still forging ahead. Perhaps one reason is that few ideas are so pro-car as this one.

However, Shared Space is really well past its "best before date" and it's time that the idea was squashed altogether rather than being the subject of pointless debates.

Still, I wish Steve Melia the best of luck with his part of this particular debate.

Jim: The Netherlands is remarkably quiet, and yes we cover this on the Study Tours. As with everything, this too is due to deliberate policy. Quite extraordinary lengths have been gone to in order to reduce the noise from roads.

It'd be great to see you on a tour next year.

Anonymous: That it becomes a train-spotting exercise to try to find exceptions is really the whole point of this article. The Netherlands is not, and never will be perfect, and I do also document the problems. However, this doesn't alter the fact that almost everyone in the Netherlands can leave their front door and cycle to any destination of their choice they'd like to cycle to, with a good degree of segregation from motor vehicles and a high degree of subjective safety for their whole journey. It's not like this is other countries.

Tallycyclist said...

Another solid post on a topic that is so often twisted out of proportion. Sometimes it may be due to genuine ignorance, but my pessimistic side feels that it's none other than purposeful misrepresentation of facts under the guise of "not knowing the truth." Several times I've come across blogs and articles making statements like "Denmark and Holland only have separated infrastructure on x percent of their roads (usually some low number like 10%)." To the everyday person who isn't a civil engineer or urban planner, the statistic sounds catchy and convincing. Most people probably aren't thinking about neighborhood streets or minor roads, yet these are all factored into the total road network. Obviously a dead-end road doesn't need a separated bike lane, but it's hard to ascertain that point simply by reading an article of the type I mentioned above. This post of yours should be mandatory for every city planner and community activist looking into transforming the infrastructure into a more people-friendly one.

Oldboy in Brussels said...

Hi David and thanlks again for this great post.

The following link might be of interest:

It's in dutch but nicely illustrated. It is a note that decribes the transport policy of the city of amsterdam, actually how Amderdam defined it's main transport networks (hoofdnetten) for the three modes: cars and HGVS, Public transport and then bikes.
The policy report is very interesting to read as it deal with quite a complex city with many old and "too narrow" streets. It really shows how, when carefully planned, it is easy to keep cars flow to reasonable level in towns to free up space for other modes and without restricting totally car acessibility.