More collisions at UK's first Dutch-style roundabout than old layout, figures reveal

I lived in the Netherlands for 3 years and cycled everywhere. These roundabouts work if driver's understand how to use them properly. Enough said.

Because you don't introduce hazards to make a road safer.

brogs wrote:

Because you don't introduce hazards to make a road safer.

Isn't that exactly the thinking behind speed bumps, bollards and road narrowing?

Mind you, the road was never really in danger.

The only hazards are the drivers and they haven't changed.

Krzystoff wrote:

Apparently speed is an issue, if so why are there no speed humps, speed cushions, or other features on the road prior to the roundabout?

NCN has tought me that if you have a speed problem approaching a junction, the best solution is a pair of fences installed to create a chicane on the approach. Why does this genius approach never get applied to roads?

Yes... I defer to David Hembrow in everything Dutch cycle infra.  I do note on this one that he's at variance with most of the Dutch also.

The usual rule is orbital roundabouts with cyclist priority in urban areas, cyclists don't have priority in less urban areas / where traffic may be faster.  BicycleDutch had an article a few years looking at this debate.  Most people, local authorities and the cycling standards authority came down on the side of "priority for cyclists in urban areas".  Only the national safety organisation and local authorities in the more rural North disagreed.  I believe (with exceptions like Assen and Groningen) it's the Dutch version of "driver country" there.

I suspect Mr. Hembrow may be technically right here.  For most people they may perceive the normal kind as "safe" - which it is, but it's possible the actual numbers show the other kind is safer.  Plus once it's "safe enough" that you and your contacts are unlikely to have a problem people will favour convenience.

It may be cyclists fear getting stuck waiting for cars (especially in denser areas where there are more cars).  People definitely like having priority (once drivers have "learned"...).  The "safest" designs are slighly less convenient for cyclists even with no cars - they're designed to slow you down by having you turn sharply.  They may take up more space also (not sure on this).

On "UK designs" - we're really good at "cargo cult" stuff where it "looks like" a Dutch design but something important has been missed.  Or the design is OK but the conditions that are required (normally traffic speed / volume / sight lines) have not been met.

Not clear on the Cambridge case but it could be that the ring for motorists in the centre is too wide / not cambered sufficiently, so they can still keep too much speed.  It may be that the legal - or upper quartile - speeds on entry are too high, or the volume of traffic is too great.  The Dutch have standards for the maximum acceptable for these also.  Above that and this design is inappropriate and another is used - or the "traffic issue" is addressed!

Finally - even though it's been 3 years it could be that it's still too novel for everyone to have "learned" this.  It may be the only one most drivers see.  Finally it's still the case most UK drivers don't cycle - even in Cambridge I imagine.  Not so in NL.

I forgot - Ranty Highwayman has done a good look at the Cambridge roundabout.  Given he's a UK civil engineer with an interest in these things it may be worth the time.  He does have videos and pictures though!  He also notes the points raised by David Hembrow vs. the Dutch standard.

It doesn't sound like the design itself is the main issue (though no mention of the camber).  I didn't notice if he had any info on traffic volumes.

Looks better than most UK roundabout designs to me.  Many of which will of course be "safe" by being so obviously forbidding no vulnerable road users go there.

If anyone's still interested there's some more interesting info and videos here:

A bit of cycle roundabout history: https://bicycledutch.wordpress.com/2011/05/12/priority-for-cyclists-on-r...

Behaviours and interactions (on annular type) - good and bad - are shown here
https://youtu.be/XhqTc_wx5EU?t=95

An example of a large vehicle having to traverse slowly (coach):
https://youtu.be/564Yf6iWjUw?t=176

His alternative, preferred, design has a major flaw because it requires cyclists to give way to traffic exiting the roundabout. This means that they're giving way to potentially every single vehicle that enters the roundabout. 

This leads to decreased use of the bike path (because why would you take the slow way), installation of beg buttons (which often don't work and will then lead to a ripple effect), or people just ignoring them.

Maybe easier to set a maximum vehicular speed of 10mph on roundabouts. Going forward limit access to cars with high EuroNCAP pedestrian safety rating so that if a driver does screw it up, the car will intervene. 

Fursty Ferret wrote:

His alternative, preferred, design has a major flaw because it requires cyclists to give way to traffic exiting the roundabout. This means that they're giving way to potentially every single vehicle that enters the roundabout. 

Correct - but it's definitely a deliberate choice rather than a flaw - you may have missed a couple of important points.

Fursty Ferret wrote:

This leads to decreased use of the bike path (because why would you take the slow way), installation of beg buttons (which often don't work and will then lead to a ripple effect), or people just ignoring them.

Citation needed!  However I would agree that this would be the likely effect in the UK if we imported these without paying attention to the following points.  The main one being "design requires a limit to the number of motor vehicles".  Of course in the UK we have that backwards - we maximise the motor vehicle capacity first then everything else comes a distant second.  Without tackling that I agree that this design would be less helpful - but if there are many motor vehicles the other design (with priority) is not suitable either.

a) With low motor vehicle numbers - as Mr. Hembrow shows in his video - it will be rare that you'll need to stop at all - even though you have to give way to traffic both entering AND exiting the roundabout.  With expected frequency you are at most likely just to slow slightly to allow a vehicle past.  This requirement is for very low (at least by UK standards) motor vehicle numbers in a given period (an hour?), not averaged over a whole day.  I think this is taken seriously, unlike the UK where we seem to set limits which we expect to be exceeded...

b) It's much better not to rely on drivers for your safety.  We all likely have experience of SMIDSY and that can happen on a straight road!  Roundabouts impose a higher mental workload on drivers - they're slowing or accellerating, checking for other traffic, preparing to turn or completing a turn.  They are simply more likely to be distracted here.  Even without satnav / passengers telling them directions / mobile phones...

With the alternate design you do not rely on motor traffic stopping.  That sounds like a regression - an inconvenience.  However when combined with "very few motor vehicles" this is actually empowering cyclists.  The "cyclist priority" annular design bakes in the assumption / requirement that motor traffic will see cyclists and will stop *.  The alternate design encodes "it's not your priority here" and so it's the responsibility of the cyclist to check if they can proceed.  Power is back in your hands.  (The design also slows the cyclist slightly with the approx 90 degree turn before crossing to encourage slower speed / more time).

Fursty Ferret wrote:

Maybe easier to set a maximum vehicular speed of 10mph on roundabouts. Going forward limit access to cars with high EuroNCAP pedestrian safety rating so that if a driver does screw it up, the car will intervene. 

Again, both designs should incorporate a) a small central section for a narrower radius and b) an adverse camber - the effect of both should be to enforce slow motorist speeds.  If you try to hoon it you'll get physical feedback telling you you're about to roll your car.  (This may be seen as unacceptably dangerous by UK road authorities of course...)  There is normally a steeper "overrun" section to allow longer vehicles to get round safely (but very slowly).

* This is a bit like unsignalised UK pedestrian crossings - the design says the vulnerable road user should proceed and the dangerous one should stop.  Now the annular cyclist priority design is pretty safe.  If at least one of the cyclist and the driver is paying attention and the annular roundabout is well-designed AND cyclists and motorist speeds are low it should be possible to stop safely if someone has made a mistake. 

I like the idea of an adverse camber.

That and adding tighter corners, engineering solutions.

Apparently smaller roundabouts (without being "ignore it" mini roundabouts) are allowed by UK standards (see this technical article on a couple of points about roundabouts by the Ranty Highwayman) BUT it's "shutting the stable door..." because the whole point of roundabouts is enhanced motor vehicle capacity over alternatives (with a reasonable level of safety - for the motor vehicles...)  You can bet there are standards to achieve here and tweaking parameters to increase speeds and vehicle throughput is still going to be priority.

I should add - the other issue with the "priority" design is that motorists may need to look in two directions for cyclists but this is not always the case.  There appear to be both single and two-way cycle paths used in this situation.  Of course if there's a single-direction cycle path it would be less convenient for Dutch cyclists who just want to turn left (same situation for a right turn in the UK) so there may be some temptation to go the wrong way.

In most (all?) of David Hembrow's examples there are two-way cycle paths so you can minimise the number of road crossings and because cars are not expected to stop for you there is no "will they remember to look both directions".  It's simply your job to check for cars - in one direction at a time (there's a central refuge between lanes).

Have you lived in the Netherlands?

There's a Magic Roundabout in Hemel Hempstead and another one in Colchester.

I'm not aware of any major issues on any of them.

I think, partly, as when a driver looks at the road signs, they are forced to slow due to the complexity of the signs.
It soon becomes apparent though that as long as you treat each mini-roundabout as a roundabout, you can go any direction around the central one.

Hmm... Mixed about this one.  Pretty sure it's "shared space, but for cars" e.g. possibly works by making everyone a bit more wary.  (That works OK if there are no vulnerable road users in the mix - otherwise it's a pretty awful idea).

Again - devil in detail - or rather in the speed and volume of traffic.  I'm guessing there are lots of vehicles at some point otherwise why would the have this.  I suspect the traffic volume would prompt the Dutch to build a turbo-roundabout here.  That essentially means that a grade-separated level for cyclists and pedestrians is also built rather than at-grade light-controlled crossings.

Not sure I'd cycle in the big ol' concrete football pitch.  Apologies for the pedantry but if they've placed a cycle lane around the outside then even less so.  A cycle path (separate from the carriageway) would seem the minimum.  Given there appear to be multi-lane entries and exits I'm a little surprised at the low accident tally.  Or is it just "clearly not a good idea to ride across, so cyclists dismount and wait at the multi-stage pedestrian crossings"?  That may indeed be safe but it's hardly convenient.

I'm a firm believer in zebra crossings on entrances and exits of roundabouts. Seems to work in other countries, so the only thing stopping it working here is driver attitude

Morgoth985 wrote:

'one regular user told the BBC the roundabout is "very hazardous" for cyclists and pedestrians, suggesting motorists "use it at quite high speed" and "it's too much for even an experienced driver to take in".'

Perhaps these drivers, experienced or otherwise, could slow the fuck down then and they might find it easier to "take in" their surroundings.

I see the cycle lane is coloured bright red, why?

Is this so that motorists, seeing that it is red, will mysteriously slow down? Experience suggests they don't GAF what colour it is.

Maybe instead the cycle lane should be raised. Motorists would then only speed over it once, before learning that dental work and suspension repairs are really expensive.

So the blood does not show.