Not-Dutch RBT

24 posts / 0 new
Last post
Not-Dutch RBT

For those who came on the second infrastructure safari in Newcastle last year I took us through a RBT where we were trying to find a way of using a peripheral cycle track.  The design is now published for consultation so online for anyone to take a look:


Situation is a UK B-Road where traffic volumes dictate 2 lane entry and exit for motor vehicles on the north-south axis but the east-west side roads are low volume.  Solution is Toucans to cover east-west cycle movements but north-south (which is where people want to go) the arms are crossed on zebra style priority tracks on raised tables.  

The tracks are danish style hybrids either side, but on the approach to the RBT they move back away from the carriageway.

This ain't dutch but it is a solution that could be applied in a lot of UK situations.  The raised table crossings mean that geometry doesn't need to be quite as tight on the RBT.

Comments welcome, the design started life on my kitchen table with my kids colouring pencils, I'd like to think that this is a step up from some of the other stuff going on in the UK at the moment.



Tom - I believe the Dutch style of roundabout is to give cars multiple entry lanes but only one exit lane, which makes crossing the road easier to judge - I assume that's kiboshed here on the dreaded 'traffic flow' grounds?


AKA TownMouse


Thats it, the traffic volume moving north south could not cope with a single exit lane from the RBT.

As you say the dutch would probably have sent some of the traffic somewhere else allowing cycle priority at every arm, but in a situation typical of the UK our "bypass" is too far away to function as a ring road for local traffic, heavier traffic volumes, so less priority....

In this situation though the main cycle flow is in the same direction as the main traffic flow, with east west cycles being mainly local traffic and school kids, so (straight single stage) toucans on the main arms is not the end of the world.  


They will be straight single-stagers then? And not pig-penned??

AKA TownMouse


No local knowledge here, but if the traffic flow (motor and cycle) is so skewed along one axis, the question for me is why is this a roundabout at all.

Why not have signal control, with the predominant phase being straight ahead only (to prevent left-hooks) and have a much shorter  east-west and left-turn only phases? Would me a two-stage crossing for cycles turning right but given the patterns you describe, would that be the end of the world?



several site specific factors killed a signal controlled X roads, including need to cater for U turns as other RBTs are being removed.  But, the signal controlled designs would not have had a separate left turn lane, no space for this given traffic volume, so would have been far from perfect.

Katsdekker's picture

From a cycle user perspective... I use the roundabout infrequently, but regularly. It's awful, and a barrier to cycling. To raise some question on the proposal:


1) The roundabout geometry does not look tight enough, so for example: turning traffic towards the zebras / cycle crossing could be too fast. Some more necking down possible please? 2) Not sure what separation method is used for the left-turning cycle track (the bit that stays on the roundabout). Should be a full kerb to ensure not being 'clipped' by turning motor traffic I'd think. 3) The track doesn't look wide enough. Is it 1.5m (min)? Overall, I can see that right and well-meaning principles have been applied, but got lost in the detail perhaps.

One procedual question. Which consultation is this? As the chair of the local cycle campaigning group I have not seen this drawing before. It makes me worry about council communication methods (again).


Hi Kat, the drawing has been shown in paper form at a cycle forum session where newcycling were represented, but it's only gone online recently as part of a consultation entitled Gosforth Transport Improvements 2013.

The cycle tracks are 1.75m at the junction, and yes full kerb on the RBT.

the geometry of the RBT is far from tight, so valid criticism, but the cycle track crossings are on raised tables  so the design is not relying on geometry (although perhaps it could do more).

Katsdekker's picture

Could the zebras be on both sides of the cycle track? That (as well as the raised table) would slow turning drivers down when they see a well-regognised symbol. I worry particularly about the northwards traffic turning West as it is quite a popular route to get onto A1 (sigh).

I also worry about the two-laned loose geometry of the roundabout. 

(little whinge. Comms are far from perfect. Council has to get much much better at reaching out, sharing more widely, better regular updates to the cycling community beyond specialist meetings such as the CCAF Tech Group... their website is constantly redone, we lose info and data. It's a mess.)


Hi Kat,

the final design as to how the Zebra - Cycle tack combination looks will be down to the DfT.  At the moment any cycle track marking next to a zebra is against the rules, but change is iminent enough for a local authority to pursue a design and negotiate for something workable.  I've seen one picture from DfT which used elephants footprints next to a zebra, which does look similar "zebras on both sides".

Will look again at the geometry on the exits, agree that if these could be tightened as well as tables that would be preferable.


I have to agree with previous posts, if you want vehicles to be slow(er) on the exits you ideally need to tighten the exit and entry radii, relying on raised tables on the exits is leaving it a little late. Maybe start off with putting in the largest circle that you can fit within the existing kerblines - the northwestern and southeastern kerblines look like they would constrain the ICD - and see how the centrelines of the approach roads relate to this circle. Then re-allocate your lanes based on these existing centrelines.

As for the crossings themselves, since 'Signing the Way' suggested that combined ped/cycle zebras,  (tigers?), might be used (at least experimentally) then I'd say keep it simple and just go with what people know, stripes. However, this might need some additional signing on the approaches along the lines of "Give way to cyclists on exit", or somesuch, until drivers got used to the idea. This would no doubt take a determined council officer to see it through with the DfT, but 'StW' is their publication so they should at least be open to the idea. There is an unpublished TRL report carried out for TfL which looks at this combined use of zebras and IIRC endorses it, or at least can find few downsides.

pete owens

There is a very good reason why parallel zebra and cycle track crossings are not permittited - and that is they operate to completely different sets of rules.

At a zebra crossing a pedestrian gains priority only after they have stopped at the edge of the road. This means that approaching motorists only need to stop for people they see waiting to cross - something that is easy to observe while approaching the crossing. This works well for pedestrians as stopping is not a problem. 

With a cycle track crossing you need to mark priority - either to the cycle track or to the carriageway. This means that either all cyclists or all motorists will need to stop at the crossing to check whether there is anyone approaching, whether or not anyone is actually approaching. 

The only way a can think it would work would be for an extra wide zebra crossing to be marked and for cyclists to follow the pedestrian rules - ie they would have to stop (and probably dismount) - but would then have priority when crossing.


Having coped with Dutch roads for 5 years as a cyclist and a driver....  I think Pete has hit identified the key issue (the other smaller issue is to make sure that the space between the crossings and the roundabout is long enough for a car, otherwise the crossings get blocked up....).

You can't put a zebra next to a cycleway without being very clear to all three users about priorities, otherwise someone will get hit.  I don't see it mentioned here, but in NL there is a system of road markings with white triangles, to indicate who has priority.  If the pointy end of the triangle is towards you, you have to give way.  So in this circumstance, if you intended the cycle strip to have the same rules as the zebra, the edge of the cycleway facing the traffic would have a line of white trangles, indicating to the traffic to stop.   Of course not many British people would understand such a line of trangles...

3 potential solutions, from my Dutch experience:

1) I think on such a busy roundabout, the Dutch would make the zebras into pelican-type crossings, and have the lights control the cycles as well as the pedestrians.  Only you'd have to think about whether to stop all the traffic on the roundabout, so needing a lot of lights, in case everything clogged up.

2) Push the zebras further away from the roundabout, so that they are not confused with the cycle crossings

3) The only time I think it would work with the zebras where they are shown is to have a cycle-lane going round the roundabout.  That's what the Dutch would do on a quiet roundabout, but that doesnt sound like it's the case here.



Signing the Way
Traffic Signs Policy Paper
October 2011

"Issues with the current system

TSRGD has also reached its practical limits. It currently prescribes over 860 traffic signs and provides hundreds of permitted variations. This high level of regulation can be a barrier to the delivery of local transport schemes that reflect local needs. Issues have been identified in the review include:

TSRGD places an unnecessary burden on local and central Government – it does not provide sufficient flexibility for local authorities to deliver certain traffic management schemes that are introduced on a regular basis. This results in over 500 requests for special signs authorisations to the Secretary of State each year;

it is essentially reactive and does not promote innovation and creative solutions. There have been several recent initiatives that local authorities have wanted to trial to promote sustainable transport, such as permitting cycling across zebra crossings in order to provide continuous cycle facilities, that are not permitted under the current pedestrian crossing regulations (the "Zebra, Pelican and Puffin Pedestrian Crossings Regulations and General Directions 1997");"

"4.13 The "Zebra, Pelican and Puffin Pedestrian Crossings Regulations and General Directions 1997" will be updated and consolidated into TSRGD to provide a single concise set of regulations."

"The flexibility to trial innovative cycle schemes
We will also permit trials for cyclists to use zebra crossings where they could form part of continuous cycle routes and facilities only, which the current pedestrian crossings Regulations do not allow. Zebra crossings are common in low speed urban areas where cycling is likely to be more prevalent. Currently a toucan crossing must be provided for cyclists to cross."


Shared Zebra Crossing Study

"7.2 Key Findings and Recommendations
7.2.1 Cyclist and pedestrian conflicts
There were few observed conflicts between cyclists and other cyclists (0.65% of cyclists, as shown in Table 5-10 if 1686 cyclists have a total of 11 conflicts with other cyclists) and pedestrians (3% of cyclists as shown in Table 5 9), and those that did occur between vulnerable users (such as other cyclists and pedestrians) were generally of a low conflict level. The majority of cyclists (87.8% in Section 5.1.4) at the study sites rode across Zebra crossings for part or all of the crossing, most of which were not involved in any conflicts. It was found that, were specific behaviours limited through controls in the built environment, cycling across the observed Zebras in their present form is not necessarily more dangerous than pushing the cycle. These controls are listed below in the Design Imperatives subsection. Research by Trevelyan and Ginger (1989) also indicates that pedestrians and cyclists share crossings with few problems."






pete owens

But that study was not looking at whether it is sensible for cycle track crossings to be parallel to zebras (as in the Newcastle design) - or even whether zebras are an effective design for a cycle crossing, but how big  a problem the illegal use of existing zebras by cyclists causes to pedestrians. The quoted conclusion effectively says "not very much" - but if you read the actual report:

Table 5-8 on page 46 shows they observed an overall conflict rate on 8% of cycle movements. True, it is mostly low level "Discomfort" level - but that is still a pretty serious rate of conflict. If you used a zebra crossing twice a day on your commute you would end up causing a nuisance to pedestrians on nearly a weekly basis. I can't imagine the author is much of a pedestrian.


Well only 8% of conflicts being level 1 or 0, on crossings no wider than 3.5m and designed solely with peds in mind is excellent.

Having watched the TRL training videos on assessing 'conflict' in vehicles at junctions then I would suggest level 0 ("Discomfort that might be caused to a party particularly by close proximity, higher speeds or unreliable movement, therefore is subjective") is probably on the level of a ped tutting disapprovingly at a cyclist as they pass, and level 1 conflict ("Precautionary or anticipatory braking or directional change  when risk of collision is minimal") would be less than that experienced by two people meeting in a corridor and doing the passing dance (I suggest that would be a level 2 conflict, "Controlled braking or directional change to avoid collision (with ample time to perform the manoeuvre))".

pete owens

Which is the sort of logic which results in all the rubbish shared use pavements we see purporting to be cycle infrastructure. Whatever you think pedestrians ought to be able to put up with they really do not like cyclists invading their space - see this from Living Steets:

However, this is still getting away from the main issue. The problem is not whether cyclists can ride on zebra crossings without mowing down pedestrians, but whether it is remotely sensible to put a priority cycle crossing immediately adjacent to a zebra crossing when both operate to a different set of rules and require completely different behaviour from approaching motorists.


Having spent many years designing roundabouts in a UK stylee, trying to Dutchify one has been a struggle. However, I've tried to create something which,  with further modification, might work in the real world UK.

The aim was to try and get something realistically buildable. For that reason the routes on all arms are mainly on-road - kerbing costs, drainage works more so - therefore only the centre island and corner radii have been significantly re-kerbed to increase deflections and reduce speeds. The cycleway circulating the roundabout is at footway height, which means the crossings are raised 'tigers', shown as 3m wide. This also means that there will need to be ramps in the vicinity of all the cycleway/carriageway splitter islands - not ideal. All entry radii and most exit radii are 10m. Lanes for motor vehicles are all 3m wide, with some flaring at the entries and exits. Cycle lanes vary from 1.6 to over 2m depending on  the arm. The central reserves on the northern and southern arms are 2m wide - this allows a 'design bike' to wait, should drivers be reluctant to stop for them, (what are shown as central reserves on the minor eastern and western arms should be triangular splitters, unfortunately these were a legacy of my construction lines). The over-run area in the centre is a result of my concen that, because of the tighter geometry, HGVs may not be able to circulate comfortably.


There you go - my first Dutch inspired effort - tear it to shreds!

pete owens

Back to the original post... As you say, not remotely Dutch - in fact a typical UK approach in which the basic design is exclusively centered around maximising the throughput and speed of motor traffic, while pedestrians and cyclists have to somehow make do with the scraps around the edges and cross where it is least inconvenient to motors. Even if you accept the case for having 2-lane entances and exists for N-S traffic, I cant see any excuse for 2-lane approaches from E and W. I would imagine that the roundabout was there from the time when the road was the A1 so there really should be scope to challenge whether there is a genuine need for a multi-lane roundabout.

Others have mentionned the geometry - which is the key element in making continental roundabouts so much less hostile for pedestrians and cyclists. By making the central island smaller the proposed design actually makes things worse by reducing the amount of deflection required to cross the roundabout to  almost a straight line. Also the crorners for vehicles turning left from the main road have been made less tight.

The toucan crossings look to be a bit of a dogs dinner. The design doesn't seem to make its mind up whether the the cycle track is part of the footway (which is implied by the tactile paving crossing the cycle tracks heading towards the roundabout) of whether it is part of the carriageway (which is the case for the cycle tracks heading away from the roundabout and which have a stop line). I'm not sure how E-W cyclists are supposed to use these crossings.

Moving the crossings much nearer to the roundabout is a real plus for pedestrians - but if they are so concerned about the volume of traffic that they insist on the need for a multilane roundabout I can't imagine they would tollerate a single stage crossing - it would need to hold the traffic on red for long enough for granny to cross 4 traffic lanes + the central reservation + one of the cycle tracks - by which time the queue will be backing up round the roundabout.

There doesn't seem to be any way for cyclists heading to the E or W to re-merge with the carriageway. I guess they would be expected to start at the crossing then swerve left across the zebra crossing. There are also a number of sharp right-angle turns featured in the design. These are a common feature of paint on the pavement type facilities, but would be very difficult to negotiate where you are constrained between kerbs.


pete owens

I looked up this report on traffic volumes in Gateshead:

On page 51 shows the traffic volume on the Great North Road as 21000 per day - and on a declining trend. That is busy but not uncommon on single carriageway roads.

TRL  9/97 describes continental geometry roundabouts:

It indicates they should be suitable for traffic flows of 2500 per hour. This suggests that even a compact roundabout should be able to carry the volume of traffic using the road - and there is certaiunly no need for multi-lane exits.


to clarify, traffic volumes on this section of GNR are closer to 30,000 vehicles per day, so combine with traffic using the minor arms and volume is above what a conventional continental / dutch RBT design could cope with, even with two lane entry and circular carriageway.

Pete, cycletrack / zebra combos when they come will include some form of give way marking, so combine a zebra for peds and the existing priority cycle track give way arrangement which works well in a number of UK locations.

Andy, interesting and will show your sketch to the scheme designers.  it looks like you've omitted the cycle track on the SE corner, which may mean you hit the same space constraint that we did?  To find the extra 2 metres of space for that track you have to reduce the island size.

My feeling is that engineers in a UK highways dept are not quite ready to use cycle-zebras on the main arms of a road carrying this much traffic, hence the Toucans, but if it can be shown to work on the minor arms then that's a step in the right direction.  

The Newcastle design delivers quite tight geometry for left turning traffic leaving the GNR, but the small island means that Broadway E-W traffic can pretty much straight line the RBT (although it is going considerably slower to start with).



The southeast corner is very tight for space and I compromised the design and shared that space with peds (although I assumed the crossings would be tigers - so there's an element of sharing anyway).

I wouldn't want to make the central island narrower than it is as the end radii will be smaller and this combination of tight radii and straights could lead drivers to misjudge their speeds as they circulate it. Ideally I'd always tend to a roundabout and central island being circular.

There might be space on the northwest corner (it seemed quite wide when I was drawing that design up), try shifting the whole thing that way - I don't think deflection should be a problem. However, some compromise may be needed, either in widths or via sharing but it may just be possible to get proper segregation all the way round.


Maybe the highways department in Newcastle might start with something quieter. Though not ideal or as aesthetically pleasing as the Dutch originals, reducing the kerb radii and providing a separate (raised) cycleway around the periphery of a roundabout could be done in bitmac and thermoplastic paint - similar construction to the raised dome of a mini-roundabout. This may also have the advantage (at least in the engineer's eyes) of being relatively more easily removed than a full-on re-kerbing scheme, not that you want it removed, but it may give them comfort. Alternatively, there are recycled rubber kerbs and raised cushions available from the likes of Rediweld, for eample, which can be used for temporary arrangements.


sorry Andy, should also have said that I wouldn't go for a cycle zebra over a 2 lane exit as you've shown, so still stuck with a toucan.  You may not need to go as tight as 10m radii on the corners, the CROW specifies 15m where there is a crossing island and 12 without (12/8 entry).

Also to pick up on Pete's point on the Toucans, they are very definately single stage, and they are closer than a lot of UK schemes would put a crossing to a RBT.  These crossings will at times shut down the roundabout.  They are not the only single stage crossing across GNR going in as part of the scheme, there is another shown on consultation drawings further south.  [lots of schools on either side of the road and pupils crossing back and forth].

There is quite a lot of space reallocation going on to create protected cycle tracks either side, mostly through reducing general traffic lane widths which will have implications for traffic speed.

I guess to some extent it's the age old question, do you do something now that can cope (just) with existing traffic volume; or do you do nothing for now except splash around blue paint and wait for long term traffic reduction policies to take effect (which you could argue is what has been happening in London).  I guess I now know where I stand on this which is to get something done now which avoids inappropriate shared use or a dual network.

Its worth pointing out that this scheme has not yet gone through all of the design and audit stages which happen in the UK and are to put it mildly tilted in favour of providing for motor traffic.  


pete owens

OK it sounds like that would be beyond the capacity for a compact single lane roundabout (though not by a huge margin).  It looks like a good case to look at installing a turbo design - as in Bedford.



Log in or register to post comments