Saturday 28 April 2018

15 metres

This week, I have been lucky enough to have had time to think about how I would redesign a secondary type of street and it's adjacent residential areas.

It's a real project which I have no idea where it will lead to (if anything), but it is great to have some thinking and designing time.

In terms of street classification, we can use the follow;
  • Major roads - dual carriageway trunk roads or motorways
  • Primary streets - single (sometimes dual) carriageway A-roads
  • Secondary streets - single carriageway B-roads or unclassified roads
  • Local streets - residential streets
The definitions are a bit more complicated than I have set out, but in essence I am interested in the secondary street. It is a pretty bog-standard suburban street which mainly residential frontage (houses and flats). It has a parade of shops, a few individual businesses and a primary school The street carries two bus routes and it conveys traffic from one local area to another. At each end of the street, it connects to busy A-roads.

In terms of its geometry, it is generally straight and it is an almost a consistent 15 metres between the highway boundaries - 15m being the minimum width. The carriageway is 8m in width and nominally marked with 1.2m wide advisory cycle lanes and 2.8m general lanes along part of the street and general lanes with a hatched median elsewhere. 

The street carries some 14,500 vehicles per day which is very high and so it is not surprising that traffic congestion is bad twice a day. 

The footways are wide with old trees towards the kerbside - I have looked at a couple of historic photos and the street has never had a verge. The street has plenty of dropped kerbs for vehicle access and a number of side streets leading off it, all with nice tight radiuses and into filtered neighbourhoods. On street parking is restricted during the day and there are some footway parking bays (all four wheels up). 

Existing layout

Luckily the streets either side are filtered, but this does mean residents have to use this street to access their homes by car and there are cases of people using parallel routes through these areas to jump the traffic queues a little bit (leaving the secondary street at one point and rejoining a bit further up) - dealing with that is important, but I'll stick to our secondary street here.

So what is the project brief? In essence, it's about trying to make the place more liveable. This means getting some people out of their cars (especially those making short trips), making it easier to cross the road, providing protection to get people cycling, reduce casualties and collision risk and perhaps to discourage a bit of through-traffic. Simple aims, but quite difficult to achieve in terms of the usual issues of cost, resident acceptability and politics (but that's all another story).

From a technical point of view, it is in theory complete possible to strip everything out and rebuild the street layout from the ground up. This approach would allow some pretty good layouts;

One-way stepped cycle tracks with buffers from traffic

Two-way cycle track with buffer to traffic

Making the street one-way would free up loads of space, but being a bus route, it does mean looping buses onto other primary or secondary streets created a much long walk for local residents who we want to be able to get the bus for some trips;

The big issue from a technical point is that a layout such as V2 means changing an existing area of footway into carriageway which is costly in construction terms and the potential need to move utilities. It is easier to convert carriageway to cycle track.

We also have the existing trees to contend with. They are mainly mature and so any proposal to remove them will undoubtedly meet with objection (although removing them and planting new ones would be so much easier). The outcome of this means we are probably looking to work with them and having the cycle track pass on the outside of them - this does mean we lose the buffer we could have had;

It does mean that we are probably left with narrower cycle tracks as we pass each tree or groups of trees, but perhaps it's good enough for what we want to achieve. As we reach the side road junctions or driveways, we can tweak the layout to give a buffer which allows the footway and cycle track to stay at the same level and be continuous with drivers having to drive over them via ramped kerb units or little asphalt ramps within the buffer;

Junction width continuous footway/ cycle track


You will have noticed that I have selected 3m wide traffic lanes. In my experience, the bus operator will always push for wider lanes. Buses tend to be around 2.5 - 2.55m in width from what I can gather (London's New Bus for London is 2.52m for example) and so this says to me that we actually need a 20mph speed limit. As well as buses becoming rolling road blocks, the speed limit is reinforced by the lane width, no centre line, plus if we hump crossings on the street, it helps with keeping driver speed down as well as creating level crossings for pedestrians.

Humped zebra crossing and a bit of a squeeze
in the cycle track to pass the tree

As usual, there are compromises, especially in trying to squeeze a cycle track around the existing trees, but for 15 metres of highway width, I think you can get a lot of improvements with something for all modes if we assume a secondary street will always carry a level of traffic where people cycling need protection, where people walking need crossings and where we facilitate bus routes.


  1. I'd love this on Station Road Hendon! (It's an A road not a B road, but plenty of space and it's a key link from Hendon station to the rest of the town, but unpleasantly car dominated. Always good to see something showing it can be done, rather than reasons why it can't.

  2. In your last example, why not put the bit of a squeeze on the carriageway?
    With the tree so close, the chance they will disturb the surface of the cycleway on that edge is not unrealistic, so narrowing the cycleway further at that point is not a good idea. People already swerve out a bit when there's a large object close to the edge, even if the surface is fine.
    Let the cycleway make a slight swerve around the tree, but keep its full width. The slight narrowing of the carriageway serves as an additional feature helping to slow down traffic to the posted 20 km limit. Buses can still pass each other in 2.80 m lanes.

    The earlier option is better.
    Alternating 15m (v1) and 15m v3 a few times (i.e. creating slight chicanes by alternating the side of the road that gets the wider verge with the trees, every 100m or so: not too closely because of the buses) would be the best option for built-in speed control. You could then put an occasional parking spot in the wider verge, between the trees, to help get the proposal accepted, as the loss of those pavement parking spots will create a lot of protest. (Movement in and out of those parking spots will also help slow down traffic slightly, without endangering cyclists or pedestrians).
    Making the wider verge green, with new trees and clear parking spots that don't encroach on pedestrians will, together with the level zebra crossings and the extra distance between pedestrians and cars/buses (and their exhausts) because of the cyclepaths will help make the street much nicer for pedestrians as well as cyclists, and hopefully get more people to accept the changes.

    1. Yes, there is definitely scope (and probably a need) to play around with the sections along the whole length of the street.

  3. Hi

    Regarding V5 (which seems like the most agreeable solution to all parties)
    Would it not be better to have the cycleway the same level as the footway considering the lines of trees effectively delineate pedestrians and cyclists. This would provide better seperation from the carriageway.

    Depending on the maturity of the trees, do you think you would be able to reduce the 1.7m width allowance to just 0.9m considering the root structures?

    This was an interesting read. Thank you.

    1. *V4.... i meant version 4 was the most agreeable. whoops.

    2. Oh it would be at the same level or slightly stepped below the footway - it's a limitation of Streetmix!

  4. Very interesting to see the options which you came up with for a greenfield site. In practise as a number of commentators have written, the issue of trees would figure prominently in public consultation. Re options 4 and 5, I have seen a 2.8m travel lane on a bus route in Assen but do you know of any locations in London (or the UK) where travel lanes on a bus route were reduced to 2.8m? This would allow the cycle lane to be increased to 2m or alternatively the provision of a small buffer (0.2m).