The following paper was presented by John Dales, Director of Urban Movement, to the All Party Parliamentary Cycling Group’s inquiry. It relates to the evidence session on Design, held on 6th February 2013.
Before turning to the mater of design, it is worth first considering the designers. As the ‘Go Dutch’ campaigning brand demonstrates, the challenge we face in Britain is not that a range of potentially appropriate design solutions is unknown; it is that too many professionals lack the knowledge, experience, training skills and/or political support to appreciate, apply and adapt for Britain the kind of physical solutions that have proved successful in other countries.
Many of the professionals whose role is, should be or will be to improve physical conditions for cycling are, like me, trained traffic engineers and transport planners; but other disciplines are often also involved, including landscape architects and town planners. Especially in urban areas, these designers are commonly tasked with trying to ‘fit a quart into a pint pot’. The many and often competing demands on scarce street space come principally from different modes of transport, but also arise in relation to the need for space for social and leisure activities. Striking the best balance in design terms can be extremely difficult, and not just because of physical constraints. The pressure from the public and/or from politicians to favour one mode or one issue over others may be overwhelming.
With this in mind, it is possible to identify a number of reasons affecting British designers’ ability to deliver better physical provision for cycling:
1. Most have been trained in a professional environment in which, to use a well-worn phrase, ‘the car is king’. As a consequence, the default in terms of street space allocation has almost always been to ensure first the efficient flow of motor vehicles. Provision for other modes has received much lower priority.
2. The idea of allocating street capacity for a user group (‘Would-be Cyclists’) that currently constitutes a flow of zero, has typically been discounted as ‘an obvious waste of money’.
3. Where the opportunity has arisen to improve conditions for cycling, designers have tended to focus on measures that ‘make life a little easier’ for people who are already cycling.
4. Designers who do not cycle, of which there are many of course, often do not appreciate the type of facilities that would be genuinely beneficial for cycling. Hence, for example, strips of coloured surfacing that hug the gutter and are poorly enforced and maintained.
5. Designers who do cycle have often failed to appreciate the concerns and requirements of those who do not. They have therefore tended to design for the ‘vehicular cyclist’ (someone who is confident to claim their place in the middle of a lane.)
6. Road safety problems are conventionally understood only in terms of the numbers and severity of recorded injuries. The principle of ‘Road Danger Reduction’ – tackling danger at source – is relatively novel, so addressing the perceptions of danger that prevent people cycling tends to be regarded as solving a problem that doesn’t exist. (‘It ain’t broke, so why should we fix it?’)
7. British designers, whose experience is largely of British cycle facilities, generally do not have an understanding of what works in others countries, or why. When they are exposed to certain techniques or ideas from abroad, a common reaction – often partly justified by the current regulatory framework – is that ‘we couldn’t do that here’.
In terms of redesigning British streets, I would say at the outset than one of the few cause-effect relationships in road safety for which there is very clear research support, is that the lower the speeds involved in collisions generally, the lower the severity of injuries. For this reason, I consider that reducing the default urban speed limit to 20mph is a very good idea, if we are serious about reducing the number and severity of casualties on our streets. I would emphasise the word ‘default’, as opposed to the word ‘blanket’. I would also add that achieving lower speeds is not a matter simply of sticking signs up. All aspects of the layout and context of any given street should be encouraging of the limit in question.
The design techniques that will best help to reduce speeds on any given street should be based on a thorough understand of local conditions. Specific speed reduction features such as speed humps may be appropriate; and if these are used, they should have a ‘sinusoidal’ cross-section, as this slows motor vehicles while being less jarring for cycling over.
Often, however, the best design approaches for speed reduction have been shown to be those that make drivers – and indeed all users – more aware of their surroundings, and specifically of one another. These can include, for example, appropriate removal of so-called ‘guard-rails’ which can give drivers the false impression that few if any pedestrians will enter their path and so they drive with less caution than they should.
2.2 Notes on evidence
Here, I think it worth dwelling on ‘guard rails’ for a few more moments to discuss the general issue of how empirical evidence does and does not influence design. For over seven decades in Britain, railings have been deployed on British streets based on the simple presumption that trying to stop people crossing streets where engineers deem it would be unsafe will have a positive road safety effect. Worse: deploying railings has become an unthinking default ‘design solution’ almost wherever there are traffic signals. As research undertaken for the Department for Transport’s Local Transport Note (LTN) 2/09 revealed, however, there has never been any robust empirical evidence that use of railings in urban areas actually does have a positive road safety effect. Indeed, fresh research commissioned for LTN 2/09 showed that collision records for comparable sites with and without railings were very similar (slightly fewer collisions where there were no railings).
This is not just a one-off example. Research for the Manual for Streets (MfS, published in 2007 by the DfT) showed that other basic road safety design standards, such as ‘stopping sight distances’ (SSD), were based on historic data that (a) designers themselves did not know about and therefore rarely if ever questioned, and (b) was badly out of date. Given that the default SSD for a 30mph street prior to MfS was 70 metres, yet just 43m afterwards, it is another reminder of how poorly-evidenced much conventional British street design practice has been.
In 2009/10, I undertook road safety research for Transport for London on eight recently re-designed busy ‘mixed-use’ streets across London. Measures had included footway widening, carriageway narrowing, raised tables, unified surface materials, and removal of ‘signs and lines’ and railings. Overall, collision numbers and severity were found to reduce appreciably, especially for pedestrians. Similar research (by others) of streets recently redesigned on more ambitious ‘Shared Space’ principles has produced similar findings. Despite the paucity of before-and-after data, it is almost certain that lower motor traffic speeds are the main factor.
However, in my London research, the collision data for cyclists was not so encouraging, and that finally brings me to the matter of designing for safer cycling.
2.3 Cycle-friendly design
I will address this issue based on my ongoing experience of designing a 7km/4.5mile-long cycle corridor in Southampton. The corridor in question stretches from the eastern city boundary, in near-rural conditions, through the suburbs, across a windy bridge, into and through the city centre, to the city’s Central Station.
In the first place I would note that a key design principle for the work is, as it should be, route continuity. The route isn’t just for those who will cycle its whole length, of course, but the adage of a chain only being as strong as its weakest link certainly applies. Much well-intended investment can (and has been) wasted if design fails to address key problem locations: typically big junctions where motor traffic conditions are very hostile to cycling.
Another key design principle is that the route is specifically intended to encourage cycling by people who currently do not cycle (or cycle very little and locally). The design team are all experienced and confident cyclists, so to ensure we are not designing for ourselves, we have (a) undertaken detailed research into design techniques used in other countries (principally Holland and Denmark), (b) enlisted the assistance of those who are familiar with cycling in those countries, and (c) gathered a small project steering group that includes people who currently don’t or rarely cycle.
Based also on previous experience and research (largely from Holland and Denmark) as to what will get more people cycling more often, we have sought to design cycle tracks that give users the greatest possible actual protection, and sense of protection, from motor traffic. The phrase ‘greatest possible’ is, of course, a loaded one – but critical, in my view, to the success of design efforts to Get Britain Cycling. The phrase implies that there are constraints on what can be done – typically one or more of funding, physical, institutional, political and operational (protecting motor traffic capacity) – and of course there are. But where the responsible authority is genuinely committed to improving conditions for cycling, then where the ‘greatest possible’ line is drawn is in a quite different place from where more recalcitrant authorities would have it drawn.
Drawing on all the foregoing, all I simply want to explain in what follows is that the design solutions that can be implemented even in the most propitious of British circumstances, still need to respond to the opportunities and constraints presented by each location. The built contexts through which the Southampton Eastern Cycle Corridor passes vary hugely and, as a result, there is a wide range of design proposals that are being consulted on, even as this inquiry session takes place. Descriptions of some of our design responses are provided below.
· A very large roundabout junction will be converted on Dutch principles, with cycling possible on separated tracks around the perimeter (see illustration following).
· Another roundabout is to be converted to a signalised cross-roads according to design principles found, for example, in Copenhagen (see illustration following).
· Two junctions where there are currently mini-roundabouts will be raised to footway level and have other measures applied to reduce vehicle speeds. Cycle priority markings will be used.
· On one quite long section, all that can reasonably be done within physical constraints of the main corridor, maintaining minimum width standards for pavements and vehicle passage and without acquiring neighbouring residential properties, is to provide a standard-width cycle track marked in the carriageway in the uphill direction only.
· Recognising the poor quality of this ‘best possible’ design within the most direct existing highway corridor, a separate route to the north has also been designed. More circuitous, but likely to be and feel much safer, this alternative route is not just a short diversion, but stretches for over a mile and runs through a combination of parkland and quiet residential streets. It is supported by a range of minor design interventions, not just signs. Specific measures include junction treatments to reduce motor vehicle speeds and a cycle-bypass of a proposed new single lane carriageway/two-way section over a railway bridge.
· In one part of the city centre, a two-way off-carriageway cycle on just one side of the vehicle carriageway is proposed, since this avoids cycle interaction with both a roundabout and three side accesses off that roundabout.
· Elsewhere in the city centre, the proposal is to relax constraints on cycling through two parks, and to provide adequately wide paths for both cycling and walking. This diagonal route is the one currently most well-used by cyclists, even if without permission, and we consider that a more circuitous dog-leg route on the streets around the park would be little-used in practice.
· In the most rural stretch, the existing two-way single carriageway is relatively narrow and there is only a footway on one side. That footway is to be formally shared by pedestrians and cyclists and a completely new shared foot/cycle path is to be created on the other side (with almost no side accesses to interrupt cycle flow). Pedestrian volumes are and will remain very low all along this stretch. Cycling in the carriageway will, of course, remain possible.
· On the long, exposed Itchen Bridge, which has appreciable gradients on both sides there is an existing, painted in-carriageway cycle track in each direction. This affords very little protection to cyclists, especially bearing in mind the potential for the frequent high winds to blow cyclists into the path of the immediately adjacent motor traffic. The proposed solution is to raise these tracks to the level of the adjacent footways, but to mark the distinction between footway and cycleway very clearly. Downhill speeds for cycling can be very fast, and so sharing the single path would be inadvisable. However, having the footway and cycleway at the same level will mean, for example, that judicious overtaking of one cyclist by another will not mean having to cycle very close to motor traffic on the main carriageway. Similarly, single cyclists will be able to position themselves adjacent to the footway, rather than the carriageway, reducing the danger should they suddenly be blown sideways.
I will finish with two further illustrations from my own recent design experience. The first (below) is a very simple cross-section diagram from a cycle track scheme on the Old Shoreham Road in Brighton that Norman Baker MP kindly opened in May 2012. This is an excellent template for a simple protected cycle track provision where space permits.
The second illustration is provided overleaf. It is of a scheme that the London Borough of Camden has recently consulted upon and granted approval of. It shows two-way cycle tracks on each side of an otherwise one-way street, where good protection is provided by the simple, attractive and comparatively cheap expedient of installing a series of robust planters and rubber ‘armadillos’. The scheme in Royal College Street is due to be completed later this year.