Monday, 19 September 2011

Connecting a roundabout with cycle paths to a road without them

I recently had an online conversation with a planner in Bedfordshire in the UK who was designing a "Dutch style" roundabout using Dutch geometry but with no special cycle facilities. To me, this sounds rather like making a cheese sandwich without the cheese. The reason why Dutch roundabouts are designed as they are was being missed on a very fundamental level by this planner.

The reason stated not to include cycling facilities was that a short way beyond the roundabout, cyclists would have to join the road again. This is inevitable, of course, in the current situation because you can't do everything at once. However, it's not a reason why the right thing shouldn't be done at the junction which is being redeveloped.

My correspondent asked for examples. Between the explanation in this blog post and my next blog post also written for this planner (showing all the roundabouts in Assen), it should be possible to gain a reasonable idea of how Dutch roundabouts work, though obviously not as good an impression as would be gained from coming here to take a look.

In reality, the difference in experience for a cyclist using a Dutch roundabout and a British roundabout is not just the geometry. Specific cycling infrastructure is an integral part of roundabout design and that is what makes cycling so much safer and more pleasant on Dutch roundabouts. It's really not the same at all to take the geometry of Dutch roundabouts without also including the cycle facilities which go with them.

Even where there are no cycle-paths on adjoining roads, there is still an advantage to having them around a roundabout. There is a big safety advantage as well as a potential improvement in convenience to well designed junctions, including around roundabouts.

I've not found a roundabout locally with no cycle facilities at all. However, this one in a village a few kilometres North of the city provides a good example of how to link a road without cycle-path to a roundabout with cycle-path, preserving safety on the arm of the roundabout which doesn't have a cycle-path.


Note that this video includes explanatory captions which are only visible on a computer and not on a mobile device.

Some may quibble at having to give way at the crossing of the other roads joining the roundabout itself. However, consider that if I didn't give way where I do in the video, I'd otherwise have done so to the same car on the roundabout itself. In practice, most usage of this junction by cyclists is on the main West-East cycle-path which we join very briefly on our way around the roundabout. Because the road to the South serves only a dead end road with one business and a farm on it, there is rarely anything to give way to. The only way to have to give way twice is to ride around the roundabout in order to make a video. Note that designs of roundabouts on which cyclists have priority over drivers have a much worse safety record than designs like this.

The industrial area just North of the roundabout does not have segregated cycle paths, but just cycle lanes on the road. In the Netherlands, this is quite unusual. There are 29000 km of cycle-path separated from the road, but just 5500 km of on-road cycle lane. However in a location like this, with a 30 km/h speed limit on a road which mainly serves adult cyclists, this is adequate provision.

The direction travelled in the video is shown by the red arrow. The pink lines are the other cycle-path which we intersect. These provide the main West-East route for cyclists. The road to the South serves one business and a farm. It doesn't offer a through route by bike or car.
The reason why the Netherlands has a comprehensive network of good quality cycling infrastructure and the UK does not comes down to two things: They made a good start, and they've continued with the effort to make things better.

What exists here now wasn't all built at once. Rather, isolated islands of infrastructure were built and eventually they came to be joined up to make the current network.

The best approach in Britain to be to insist on building examples of the best possible infrastructure rather than taking the approach that it's not worth doing something because it isn't already done elsewhere. It is inevitable that in the UK at the moment good examples will be isolated islands. The important thing is to make sure that those isolated islands are of good quality. Good things sell themselves. Once something has been shown to work well, as Dutch roundabout design including cycle paths certainly do, then it is easier to make a case for more of the same elsewhere.

Update November 2014
I deliberately did not name "the planner" who inspired this blog post more than three years ago because my communication with him had been anonymous. Our correspondence was entirely online and unfortunately was rather repetitive because he was struggling with the concepts of Dutch roundabout design. Many emails were sent, many examples given and I made the suggestion that he should come and take a look at real Dutch infrastructure before going ahead with building the flawed design which he was proposing.

Three years passed. That's more than enough time to have thought through and improved the original design. More than enough time to have come to find out how the Dutch really design roundabouts. Unfortunately, neither of these things happened. The same planner has recently designed and got built a roundabout for Perne Road in Cambridge which is almost exactly as he'd suggested he would build in Bedfordshire three years ago. It took just a few days for this dangerous design to cause injuries.

"The planner" actually boasting about his design online, so I see no reason why I should not now name him as Alasdair Massie. This blog post and the the following one showing how all the roundabouts in Assen have cycling infrastructure, were both written as part of my attempt to provide Alasdair with a free education about how Dutch roundabouts are really designed. I wanted to help him not to make the mistake that he has made, to try to avoid people being hurt by a bad design. Unfortunately, some people simply won't listen to advice.

More recently I have described how the safest Dutch roundabouts are designed. This is the blueprint for how roundabouts in other countries should be designed.

We are still available to show planners from the UK, or other countries, what true best practice looks like.

Please also read about other real Dutch roundabout designs

We offer Study Tours in the Netherlands in order to provide a service to campaigners, planners and other parties with an interest in how the Netherlands has achieved its extraordinary modal share and safety figures for cycling. Uniquely, we've experience of both the UK and the Netherlands and know how the two countries compare from a cyclists point of view. We offer a time-saving way of finding out what the important differences are, rather than having people making guesses from the other side of the North Sea. Thus far, no-one responsible for designing cycle provision in the UK has come on any of our tours.

5 comments:

Neil said...

That very neatly echo's a realisation I had just come to.

I've just been to NL on holiday and managed a few cycle outings - Haarlem/Heemsteede and Wassenaar/Leiden (and the Duines). It often wasn't up to the level that Assen seems to have. Much more cycle lanes than perhaps I expected. BUT, I realised as we came back that I think all the traffic light junctions did have at least some separation. And there were many small/medium roundabouts, but IIRC, all had separate cycle paths.

i.e. I can't recall any situation where the infrastructure disappeared or got significantly worse at the junctions.

I've always thought that junctions are critical, but the revelation from my Dutch cycling is that they do not seem to compromise at junctions. So I was just formulating a theory that if we could at least get the junctions sorted out in the UK, without worrying about lanes or paths, then that would be a huge part of the problem. And now I see this post of David's which seems at least in part to confirm that.

David Hembrow said...

Neil, junctions are crucial. However, I think you're making 2+2 equal five: The cycle-paths between the junctions are important as well. This post is not intended to suggest otherwise.

The inspiration for this post was a planner in the UK who told me he was designing a "Dutch style" roundabout without any cycle paths. This is such a fundamental error. Without cycle paths it's not a Dutch style roundabout.

It took us many holidays here to decide to move to the country. It then took us a lot longer to get an appreciation of what works and why things are as they are. This is what I try to distill into the blog and also the study tours that we run for planners and campaigners.

We do our tours in this area for two reasons. First: it's where we live and we know it very well. Second: the infrastructure here is very good, even by Dutch standards. That of course also fed into our decision to move here.

I'm afraid that the area you were in doesn't have the best infrastructure.

Neil said...

No it definitely wasn't as good as you describe and it was noticeable when you were on better infrastructure, but it was a revealing that the lesser cycle lanes nearly always had something extra done at junctions (rather then the UK approach of giving up).

I think the inferior cycle lanes with proper junction we cycled on were infinitely preferable to most UK infrastructure. And my feeling that in part this was due to concentrating on the critical part - the junction.

BTW - I meant "without worrying about whether to use lanes or paths".

MattP said...

The CEN comments rantings over the child cyclist being hit by a driver on the Radegund Road roundabout is quite upsetting.

I posted some observations which I thought would unite people to tackle the councillors and local authority who messed this up.

Instead I'm flamed with 'cycle lobbyist nutter' etc.

Although the demographic on such sites is very narrow, I do wonder if the reason we have such crap cycle infrastructure in the UK, is because it is from this 'anger towards cyclists' culture that councillors and engineers emerge.

Multiparty Democracy Today said...

Looking at this from a teenager's perspective, it is little wonder that many point out so often the hypocrisy of adults. Do these adults seriously think that they could command respect when they do foolish plots when they themselves pride themselves on being well educated and over the age of 25 (which not coincidentally is usually the age at which one may rent a car).

If these planners want to set a good example, they must pay attention to the input of people who actually know a lot about the topic they talk about. You David, you Mark Treasure, etc. We need to get beyond hype. A law prohibiting advertising government plans based on false or twisted or out of context imagery or using things like "longest X in X place" or something similar, would hopefully make hype less of a factor, though people will always try. Experiments like the rain sensor in Groningen you talked about before could be re purposed to give more and more green time to cyclists, and it has been proven now that the hype is not needed, as the traffic hasn't been delayed when rain comes down in buckets.

I'd like to see "Going Dutch" be a trademarked phrase that only a group of traffic experts with as much or more experience as you David, could allow the use of the phrase if it really would be a thing that the Dutch would encounter on an everyday basis and uses the correct context, taking into account the failures and successes of various schemes in the Netherlands. I wish it didn't need to be trademarked, but I guess with how Britain and other places in the EEA/EU have used it, it needs to be protected or it risks downplaying everything the Dutch have done right.

Good blog post about the transitions. The transitions I usually see from sidepaths to roads are like wheelchair ramps, not designed for cyclists in mind, often with an upstand to go over a couple centimetres high. Bike lanes going into normal lanes of traffic are clearly not being used in the right places or using the right designs. In the Netherlands they would only be used on low volume 30 and 60 km/h roads (urban and rural respectively) after turning the corner, if the gateway design is not used. The road traffic here is often at 50 km/h, sometimes 60, if you're lucky, 40, and in high volumes (relatively speaking, still low volume enough to be a single lane/direction distributor, but too high for safe mixing even if the speed was low). The transition provides no obvious reason for why there should be such a transition, and there is no taper. It just ends, literally that is what it says on the sign and asphalt markings. Nothing to slow the cars or make them understand that the cycle lane is ending. They couldn't care more than the guy who invented a literal flying f"@k.