Thursday, 1 May 2014

Sustrans Handbook for cycle friendly design - a poor design manual which sets very low standards

Not only does this Sustrans route consist
of nothing but loose pebbles, there's a
gate on it which I could not pass without
removing my bike trailer.
Sustrans (the name means sustainable transport) is a high profile campaigning organisation in the UK. They have a long history, having been around for 37 years. It might seem surprising that an organisation which claims to have been working on behalf of cyclists for so long should be criticised by many cyclists, but it's very common to hear their work criticised and there are good reasons why. The Sustrans name is not associated with high quality infrastructure.

A common problem with Sustrans
signs is that they tell you the same
thing in more than one direction
Several years have now passed since Sustrans proudly announced that they had built up a 10000 mile network. Unfortunately, those of us who have tried to ride along it know that this in many cases this achievement came about as a result of prioritizing quantity over quality. It makes for a good headline, and this perhaps works well for fund-raising, but the result doesn't make for good cycling. When I lived in the UK, I tried to use Sustrans' network of infrastructure myself and was disappointed more often than not. When I rode from Land's End to John o'Groats, I tried on several occasions to use the National Cycling Network infrastructure, and each time it caused a problem. In the worst case it nearly caused me to crash so at that point I switched to a dual carriageway A-road for my own safety. This is not how it should be.

Sustrans' attempts at building infrastructure often include such features as paths shared with pedestrians, on-road cycle-lanes shared with buses, narrow widths, rough surfaces, places where it's necessary to dismount, and many obstructions. These make progress by bicycle slow (so much for a practical means of transport). Where Sustrans routes consist of nothing but sign posts on country lanes, these almost invariably result in a long detour. These are not characteristics of proper cycling infrastructure.
A photo from nearly ten years ago. From this point, turn left and take the Sustrans NCN route along country lanes to Newmarket, 15 miles. Turn right to take the direct route, just 6 miles. This has still not been improved.
Fast forward to today
I could spend a lot of time writing about past errors of Sustrans but this blog post is not about those. What I want to write about today is the new Sustrans "Handbook for cycle friendly design", which sadly is not nearly the document that it should be.

Cycle training cannot
result in mass cycling
Sustrans' document starts with a foreword which includes sort of text that we've come to expect from any document about cycling. There are many nice words about "cleaner healthier travel", "public health" and "liveability" and the introduction ends by talking about doing "all that we can to enable travel to be both healthier and better for the environment".

A few pages later the document starts to talk about "understanding user needs" with the illustration to the right being prominent. This shows "primary and secondary riding positions", primary meaning of course taking up a position in front of motor vehicles in order to control the following traffic.

Sustrans say they are designing for "the less confident cyclist" but this apparently means "a sensible 12 year old [...] trained to Bikeability level 2". While Dutch children have cycled to school as well as for other purposes for many years before they reach the age of 12, the needs of younger British children to be able to do the same are not being addressed.

It should be noted that Bikeability training in the UK is not associated with a higher rate of cycling. Training cannot result in true mass cycling because it does not address the fundamental issues which face cyclists. In particular it does not address subjective safety issues so parents won't allow their children to cycle even if they have been trained.

The importance of a fine grid of infrastructure appears to have made it into the Sustrans manual (they call it a mesh) and it's illustrated by a map which similar to those you'd see in the Netherlands, showing a primary grid which is linked by a secondary and third level grid. The language of what is required is in the Sustrans document, but what follows falls well short of the extremely high quality and convenience required in order to attract people to cycle.

Sustrans emphasize Shared Space
features early in the document
The first section on street design emphasises Shared Space designs, though in the Netherlands these have been found to be dangerous and to discourage less confident cyclists.

Bizarrely, their diagram to illustrate "good street design" shows that roads on which there was space for a cycle lanes should lose those lanes as they approach a roundabout which itself has no clear markings. This will generate conflict and danger just as it does at similar designs which have been tried in the Netherlands.

Two out of three are dangerous
The Sustrans design for "road humps" and "speed cushions" has proven to be dangerous in that drivers often pull sharply to the left before reaching such a traffic calming feature in order to avoid all four wheels going over the traffic calming feature. By doing this, they often pull into the path of cyclists.

Similarly, the central island design is lethal. Sustrans suggest painting bicycle symbols on the road, but this paint will not prevent drivers from pulling into the path of cyclists in order to pass the pinch point. We have pinch-points like this in the Netherlands, but not with cyclists are on the road. They are used at entrances to villages and to assist people to cross the road.

Advanced stop lines ? In 2014 ?
We move on to Advanced Stop Lines (Bike Boxes). This infrastructure claims to give cyclists a head start at traffic lights, but they're associated with cyclists stuck in the feeder lane when the light goes green being injured by moving vehicles, especially when they turn across their path. These are very much out of fashion in the Netherlands and those which still exist are relics of the 1980s. We have none remaining here in Assen. ASLs are no longer built in the Netherlands because they are an inferior way of designing a traffic light junction. We have no ASLs left in Assen and I have long recommended that these are one of those examples of infrastructure which should not be copied.

It's already proven to be dangerous
in Southampton and lethal in
Denmark. Why duplicate this
elsewhere in the UK. There are
better, safer designs for traffic
light junctions.
Under the banner of "innovative cycle facilities" we see "hybrid cycle tracks" which sadly found their way to the UK as a result of a misunderstanding when visitors from Cambridge saw a single very old cycle-path in Groningen and though it looked achievable. This is not something to copy. We also see "armadillos", much criticised in London and barely resembling the very oldest infrastructure here in Assen.

But the worst of their "innovative" facilities by some margin is the much criticised Southampton two-stage turn. This attracted criticism well before it was built because it was no more than a poor copy of a type of junction which is not only inefficient for cyclists to use but which has also proven to be lethal. Denmark has many junctions of a similar type and has worked for years to try to make them safer. Nevertheless, a relatively developed form of this type of junction killed seven Copenhageners last year. Is that the type of innovation that the UK needs ?

Sustrans asks for British cyclists to have
cycle-lanes narrower than the Dutch build
Sustrans then moves on to on-road cycle-lanes. The Netherlands has few on-road lanes because of the many problems which they cause. I've documented those problems together with the required widths in the Netherlands. I don't suppose it will surprise many readers to find out that Sustrans has much lower standards than do the Dutch.

Sustrans propose mere 2 m wide cycle-lanes as being adequate even with traffic flowing at 40 mph. At these speeds, proper high-quality segregated paths are required.

Gilbert Road Cambridge is used as a
"good" example by Sustrans. Actually
it was a missed opportunity to do
something better.
I was also amused to find that Sustrans had includes a photo of Gilbert Road in Cambridge as an example of good design. I know this road well. It was part of my commuting route for many years when we lived in Cambridge and the rework which has been there is to a far lower standard than was possible. Gilbert Road in Cambridge is a prime example of where a very much better solution was possible but cyclists ended up with nothing more than narrow on-road lanes largely as a result of not understanding what was possible and having low ambitions for improvement.

What's more, there is the suggestion that building a cycle-lane between lanes for motor vehicles is a good idea. This is one of several points in the Sustrans manual which I've criticised on several occasions before, including in my recent summary of cycle-lane problems. This dangerous idea keeps being proposed in inferior design guides from around the world. A cycle-lane like this almost invites motorists to turn across the path of cyclists and it is not a safe place to be. Not for an experienced cyclist and also not for the 12 year old with Bikeability training which Sustrans claim to be designing for. There are far better, far more convenient designs of traffic light junctions than those which Sustrans seems to be aware of.

Buses and bikes should never be mixed
Sustrans appears to think that buses and bikes can share lanes satisfactorily. This is a combination which never works well for reasons which any cyclist and any bus driver will understand. While the average speed of buses and bikes across town may be similar, cycles travel at their average speed while buses make that average by travelling relatively quickly between stops. The result is that bikes hold up buses and buses hold up bikes. Quite apart from the potential for lethal results from genuine error where you have a very heavy and large vehicle sharing a lane with a vulnerable cyclist, the tensions that result from this "sharing" often bubble over into deliberate violence. Just two years ago an incident due to an angry bus driver in Bristol, the home of Sustrans, made the national news in the UK. This conflict must be designed out of city streets, not encouraged by bad design.

The only way to achieve harmony between buses and bikes is to keep them apart.

Cyclists need all the help they can get when approaching and
negotiating a roundabout. Stopping the cycle-lane early so that
cyclists can "mix with traffic" is not a solution. Click to find
out how the Dutch build safe roundabouts
.
The roundabout designs proposed by Sustrans are also compromised. There is great emphasis on "continental design" and emphasis on changing the geometry of the roundabout. In fact, what keeps Dutch cyclists safe on roundabouts is not the geometry, which varies considerably, but that bicycles are not ridden in the main traffic lanes.

i.e. The very safest designs of roundabouts for cyclists are safe precisely because cyclists don't use the roundabout.

Sustrans are not the first to make this mistake. It's a misunderstanding which has come up repeatedly with British road designers. I had a prolonged online conversation with a planner in Bedford in 2011 about this exact misconception. Regardless of this, Bedford has more recently gone one 'better' and proposed a turbo-roundabout with on-road cycling, and Sustrans amongst other campaigning organisations actually approved of this as good practice.

Turbo roundabouts are absolutely not for cyclists to ride around. They are a special design of motor vehicle specific roundabout intended to speed up traffic around such places as motorway exits. Ideally, cyclists won't even see these types of road junctions.

Mini-roundabouts give little reaction
time and can be more dangerous than
full sized roundabouts for cyclists
Mini-roundabouts are also featured in the Sustrans guide. I know I'm not alone in finding these to be less than safe. Earlier this year, my mother was injured cycling across a mini roundabout in the UK, just a few miles away from Sustrans HQ in Bristol. I've also had incidents in the past, and there were fatalities at mini roundabouts near where we lived in Cambridge. These are not good cycling infrastructure.

No. Don't do this.
When we come to the discussion of segregated infrastructure, Sustrans' old problems come to the fore. As I've tried to explain in the past, bollards should be used sparingly. The last photo at that link shows how the defunct Cycling England had much the same ideas about good practice as does Sustrans, but this not only costs more than would a single bollard which was adequate to stop people from driving cars along here, it also is dangerous for cyclists and restricts access by people with tricycles, cargo bicycles, trailers, or by people with disabilities. It is that latter group who I think need to be thought of most, and I'll come to this again in the conclusion below.

What Sustrans thinks of as an off-road facility is interesting. Note that we don't have any shared use paths in the Netherlands because they cause conflict and are not efficient to use but Sustrans expresses a strong preference for shared use on the grounds that it "maximises the usable width". As we're on the subject of widths, what do they suggest ? It turns out that a 3 m wide path is considered to be adequate for a main cycling route. That is thought to be enough for two way cycling combined with 2 way walking. 2.5 m and even 2 m widths are also considered to be adequate in some situations.

Sustrans recommend just a 2.5 m
wide bidirectional cycle-path
through underpasses. This is
very narrow. They also suggest
long subways are permissible.
All the dimensions are below
Dutch standards for tunnels.
There's also little consideration
of social safety.
A two metre wide path is simply not wide enough to allow for safe use in both directions at once, even without pedestrians also using the path.

Contrast this with the situation where we live in the Netherlands. We have mostly 4 m wide cycle-paths for bidirectional use, narrowing sometimes to 3 m wide for secondary routes. These are usually parallel with a 1.5 to 2 m wide path for pedestrians. For single direction cycle-paths, 2.5 m wide is normal, again parallel with a 1.5 m to 2 m wide path for pedestrians. The usable width is not being "maximised" by Sustrans' guidelines, but actually it is being set very narrow indeed.

Sustrans dimensions for segregated
paths would be OK if this was for
single direction use. A 2 m width for
bidirectional use is just not enough.
But wait, they're also talking about equestrians using the same paths. "Greater width" is then required. But horses should not be on the same paths as cyclists and pedestrians for reasons other than width. Horses are scared easily and that can lead to conflict and danger, and they leave behind something that no pedestrian wants to stand in and no cyclist wants to ride over. Horses need separate paths from cyclists and cyclists need separate paths from pedestrians. That's what we have here in the Netherlands.

Sustrans have been building inadequately wide infrastructure for an overly broad user-group for many years and of course they have seen the conflicts that result from cramming people onto narrow badly designed infrastructure. This document even suggests a way of trying to deal with it: 'On unsegregated paths consideration should be given to the erection of courtesy signs such as “cyclists give way to pedestrians” or “share with care”.' Needless to say, well designed paths do not need such signs.

Then there are the signs. Some of those shown as good examples  also are simply not good enough. The one on the left is hard to spot (from experience) and doesn't actually tell you where you're going or why you should follow that arrow. Perhaps your destination is in the opposite direction.

The one on the right illustrates another problem. This road junction has been designed in a way that it is so confusing to use that people require a sign just to tell people how to negotiate the junction. It doesn't tell the user anything about their destination unless that happens to be "Canal" or "Ashton Road".

And I've even pictures two of the others, which say "Please give way to pedestrians in path", raising the question of whether this would be necessary if the infrastructure had been designed to remove conflict between user groups instead of creating it, "Use diversion when route ahead flooded", which raises the questions of why there is a cycle path which floods and whether this is really an issue which should be "solved" by putting up a sign.

Who are Sustrans designing for ?
I mentioned briefly above that people with physical disabilities particularly have an enormous amount to gain from a real grid of high quality infrastructure. But that, sadly, is not what Sustrans are planning to enable. They've adopted the language of providing a real grid but their standards aim far too low to provide infrastructure which can be used by the entire population.

Sustrans labelled this photo
"inadequate drainage". I'm
more interested in how you
get through that gap with
this type of bicycle.
By concentrating on their mythical able-bodied 12 year old with Bikeability training, who they think can negotiate all types of roundabouts from mini to turbo, who can avoid being crushed at two stage turns, swerves easily around bollards, is un-phased by cycling in a narrow cycle-lane alongside 40 mph traffic, or indeed a lane with traffic on both sides of his lane, who has no concerns about social safety, likes sharing space with buses, pedestrians and cars as well as with horses, who doesn't mind routes being indirect and inefficient and has no problem at all with having to stop and read a sign just to find out how to cross one junction, they've seemingly forgotten about everyone else.

The designs that Sustrans are promoting are not good enough to get the masses to ride bicycles. They're also not good enough for confident cyclists to make the efficient journeys which everyone who cycles wants to be able to make.

Cycling should be for everyone. Able-bodied, disabled, young, old, fast or slow. Infrastructure for cycling should be designed for all these people to use safely at once. There should never be a choice between a safe option and a fast option. This is not a dream, it's reality just over here on the other side of the North Sea.

I think it's notable that while the best practice in cycling is not to be found in the UK, the references section of Sustrans new handbook includes only UK sources. Sustrans' references section tells me where to find their own publication about dealing with Japanese Knotweed, but there's no reference to the world's best practice cycling design documentation from CROW.

It's not just CROW that they're ignoring. We've offered to help to educate Sustrans planners about best practice but ten years have gone by without any interest being shown by Sustrans. We're still here. We can still help. We can show you better than ever what best practice really looks like as well as explaining the pitfalls that you must avoid. But we can only do so if you talk to us.

Other new guidance
Several of my older photos are used
in the camcycle guidance, one from
2008 fills most of the front cover.
The document could have benefited
from newer ideas and newer photos
but these were not sought.
As well as the Sustrans handbook, two other documents have appeared in the UK which seek to offer guidance on cycle facility design. Making Space for Cycling produced mainly by Cambridge Cycling Campaign and Space for Cycling from CTC. Both of these are much shorter documents than that from Sustrans and neither of them are so prescriptive as Sustrans' handbook. Both of these other guides fall short mostly by omission so I see them as much less harmful than the Sustrans handbook.

Both CTC and Camcycle have used photos from and around Assen, some of the photos used even feature Judy and myself. This should not be taken to indicate our endorsement. While both of these are improvements over Sustrans' work, neither of them aims quite high enough. For instance, some of the photos of Assen used as good examples are of infrastructure replaced years ago, there are references to hybrid cycle lanes and requirements for widths are inadequate.

It's also worth pointing out that not one of the three new documents discusses the safest roundabouts or the safest traffic light designs from the Netherlands. In all three cases they are promoting inferior designs which are less convenient as well as less safe for cyclists. Read more about good junction design.

From "Fast Forward" onwards, the illustrations and photos above come from Sustrans' handbook except for those on the cover of Making Space for Cycling, the largest of which is my photo.

A reader confirmed that the "inadequate drainage" barrier still exists and pointed out where it is.

16 comments:

Neil said...

The bollard/gap problem is they seem to be intent to keep out motorcycles. Even though that is impossible without excluding many cyclists.

jrg said...

Yes! worse, it won't keep out an anti-social motorcyclist if they are determined to break the law, but it will keep out many cyclists (trailers, cargo bikes, trikes, etc)

haya said...

Many of the mistakes in Sustrans' design manual are also seen in the Japanese counterpart, "Guideline for creating safe and comfort bicycle environments" by Ministry of Land, Infrastructure, Transport and Tourism. (https://www.mlit.go.jp/road/road/bicycle/pdf/guideline.pdf)

Cycling in Japan is currently popular relative to UK probably because riding a bike on sidewalk is accepted. But it have been causing conflicts between bikes and pedestrians, restraining the country's average trip distance by bike (0.75km/person/day).

To solve the problems, experts learnt from "leading bicycle countries" and formulated the guideline in Nov. 2012. Unfortunately, the focus of this guideline was to throw cyclists out of sidewalk and force them to obey the original traffic law (= ride on road) not considering their perceived safety.

But almost all cyclists here in Japan, including very influential figures, welcomed this flawed guideline and have been insisting that authorities and communities should build roads accordingly.

One of the experts who involved in making the guideline explicitly cited "Design manual for bicycle traffic" by CROW several times in his papers, but it seems he didn't correctly understand it.

Berno said...

I've been supporting Sustrans for many years now. While always accepting that they did no more than design and build touristy cycling routes, I never fully realised how limited their outlook really is. That they have been ignoring your offers of showing them examples of best practice and pitfalls in The Netherlands is unforgivable. Why is it that British engineers apparently can not accept that they could learn something from foreigners? I get the same attitude from them when I try to explain the benefits and advantages of the Dutch hybride bicycles over their drop handle tourers in the drive to get more people cycling. Thanks David for your thorough criticism, I think I’ll have to reconsider my Sustrans membership.

Andrew K said...

It should be renamed the "Sustrans Handbook for creating conflict between cyclists, motorists and pedestrians".

Al said...

with Gilbert Road, Cambridge, the issue was political will. There was a push for separated cycle lanes (admittedly, not a public one) but it rapidly became clear there wasn't a chance of that happening. Just getting on-street parking removed (a pretty useful step) was done in the face of very fierce opposition.

But yes, Sustrans are basically beyond a joke.

Paul Cooke said...

Sustrans NCN41 between Gloucester and Cheltenham... direct it is not and it's littered with barriers that block me and my trailer and definitely block disabled, tandems, trikes and recumbents...

http://www.sustrans.org.uk/ncn/map/route/route-41

bjh21 said...

NCN 51 isn't quite as bad as the signs say. According to Google Maps, it's less than 13 miles from Bottisham to Newmarket on NCN 51, and and over seven miles by the direct road. This is still pretty terrible, of course.

Jason Torrance said...

Hi, I'm the Policy Director at Sustrans. I've worked for Sustrans for some five years... but involved in campaigning for sustainable transport since the early 1990's. In terms of European best practice I and others at Sustrans consider this crucial.

One of our latest projects seeks to replicate the Copenhagen Bicycle Account http://t.co/ojAHA7Wa3v in seven UK cities... and a lot of visits to cities engaged in "leading bicycle coutries" have taken place... both to inform this project - but many other of our projects and our work more generally.

In terms of The Netherlands.... our team developing plans for a London Quietways Network http://www.sustrans.org.uk/news/sustrans-hails-big-step-making-cycling-normal-london recently had a extremely
productive trip organised by the Dutch Cycling Embassy.

On a wider front we're working with European cities to reduce car use http://www.sustrans.org.uk/news/sustrans-advises-european-cities-encouraging-active-travel.

... all small parts of working towards our wider vision where 4 out of 5 local journeys can be made on foot, by bike, or by public transport. http://www.sustrans.org.uk/about-us/our-vision

Thanks for your offer of 'educating' our design and engineering teams.... who are experts in their own rights.... how would you like to take a conversation forward?

David Hembrow said...

Jason, it's good of you write. I thank you for this. However, I find what you have written to be a bit bizarre. All four of the links that you provide lead to web pages which are about publicity and marketing for and by Sustrans. At these links I read about such things as calls for action, "ambitious plans", training and personal travel planning.

Not one of these links leads to anything which suggests a better understanding of what it will take to truly transform Britain into a "cycle friendly" nation than is displayed by your recently published handbook, which I criticized above because it is simply not nearly good enough.

While you may have worked with many different people and some of your staff may have enjoyed their trip to the Netherlands, it doesn't seem to be the case that any of this has rubbed off on Sustrans and improved the standards towards which you are working. Your new handbook only reinforces the idea that Sustrans does not understand what is required to make cycling easy to do, useful, safe, convenient and accessible to the masses.

You mock us by putting the word 'educating' inside quotes, but whether you like it or not, Sustrans really does need educating. We're offering something that you could really benefit from and we'd be very pleased to talk to your people and show them what best practice really is. We're not at all difficult to find.

fred said...

Jason - it's not a design and engineering problem that Sustrans is facing - I'm sure the designers and engineers you work with know their business - it's a policy problem. Your policy right now is 'something is always better than nothing' - and the problem with this is that it provides cover for local authorities to do the minimum possible to make a cycle route - and makes for routes that are essentially useless, cutting through muddy fields, and following very busy roads. You need to replace this with a policy of 'Good enough for all users, or not at all..'. Everything that you label a cycle route has to be usable by families,by children, by elderly people, by hand-cycle and tricycle users. Right now, your engineers and designers haven't been designing to this standard because your policy hasn't asked them to. Change the policy, and then work with them to make it happen.

Jitensha Oni said...

Jason,

I can hardly wait for your organization to upgrade parts of NCN4 in line with your handbook. I'm particularly excited by the prospect of the width guidelines being applied to the stretch north of Kingston upon Thames to Teddington Lock, starting at

https://www.google.co.uk/maps/@51.424191,-0.306743,3a,19.8y,329.09h,86.08t/data=!3m4!1e1!3m2!1swiwmQAC6Qg45TKR1JcC0qA!2e0

and the width and surfacing guidelines on the NCN4 Thames Path in Surrey between Hampton Court and Weybridge, as at:

https://www.google.co.uk/maps/@51.384171,-0.433031,3a,75y,303.84h,60.53t/data=!3m4!1e1!3m2!1slFkIeP0W7cdZ2HOGVXuevg!2e0

Please let us know when you expect this to be completed

J.O.

Mark Nokkert said...

Thanks for this excellent post. As a Dutchman trying to negotiate the British roadsytem on a daily basis, I have become exasperated by the lack of willingness from British planners to learn from good examples and simple design solutions such as abound in the Netherlands. Keep pushing the buttons, David _ one day, they (including Sustrans) will wake up and start doing the right thing here as well; Mark.

Kevin Love said...

Is there any real reason why organizations like Sustrans or Ontario's Ministry of Transportation cannot simple adopt the CROW standard. Why re-invent the wheel?

The government of The Netherlands has even translated CROW into English. What could be more convenient that simply taking it off the shelf and using it?

Kevin Love said...

I burst out laughing at the probably unintended humour on page 13 of the Sustrans manual. Take a look at the photograph captioned "Advanced stop line with feeder lane, London."

Yes, there is a car driver blocking the bike box.

Meaning that their own photograph vividly demonstrates why ASLs are a bad idea and do not work.

Alan Braggins said...

I assume Jason has put "educating" in quotes because he is aware that Sustrans have no intention of ever learning anything. I think the best way forward would be for them to change their name, freeing the "Sustrans" name for use by a genuine sustainable transport campaign.
(On the subject of Sustrans signing, I used a commuting route that follows much of NCN51 for some time before spotting the sign in this photo: http://www.cyclestreets.net/location/10359/ (Once I had seen it, I continued to ignore the detour anyway.))