Friday 17 January 2014

Inadequate infrastructure causes injuries. Better infrastructure prevents them. Learning from two minor crashes. Mini Roundabouts are unsafe for cyclists

My mother lives in Burnham-on-Sea in Somerset, England. Just before Christmas, my mum was involved in a crash while cycling. She had right of way when turning right on a mini roundabout. A driver coming from her left drove into her without seeing her.

My mother was following the line shown in red. The crash happened at the blue cross. See it on Google Maps.
The driver who hit my mother is not a bad person. She was apologetic from the start, took my mother to hospital and has checked up on her since then. The driver also paid for repairs to the bicycle and admitted liability to the police.

My mother was not seriously hurt, but her minor injuries have caused pain for some weeks now. A couple of days ago, my Mum collected her bike from the bike shop which made the repairs and rode it home. That's brave. Many people never cycle again once something like this has happened to them. While careful cyclists rarely have collisions anywhere in the world, including the UK, cycling becomes subjectively far less safe for people who have had this type of experience.

Crashes like this are quite common in the UK. They're usually referred to using the word "SMIDSY" ("Sorry Mate, I Didn't See You") as that's what drivers say so very often to the people who they have injured.

However, while crashes like this are common, they are rarely the result of deliberately dangerous behaviour. This is why neither legislation against drivers nor additional training are likely to make such events less frequent.
Another view of the same location. The red line shows the route that my mother takes to ride her bike home from the town centre.
The Blame Game
Who should we blame for this crash having occurred ?

My mother is a very careful cyclist and was riding according to the highway code when the crash occurred so she can't be blamed. The driver of the car has to take some responsibility because it was her action which caused the crash.

However it would be a mistake to think that these two parties are the only people involved. We need to look at why the driver made the mistake. This didn't happen for no reason at all, it was at least in part the result of the environment. Because crashes like this continue to occur I don't find it constructive to continue to blame either of the two principle actors.

It's not just Burnham. Britain has many
mini-roundabouts. This one near our old
home in Cambridge often caused us
problems by bike. Car crashes also
occurred. At least one person has died
at this junction since we left the UK.
This is a busy road junction yet it's designed as a mini-roundabout. Mini-roundabouts like this can be quite efficient for motorists but many cyclists find them to be dangerous. On a mini-roundabout, people have far less time in which to make a decision than they do on a full-sized roundabout. This leads mistakes being made and makes it far easier for motorists simply not to notice bikes and their riders amongst the more common and far larger motorized vehicles.

The speed limit at this junction is 30 mph (50 km/h). Burnham doesn't have any lower speed limits than this, not in residential areas which should not be used as through routes by car nor even by schools, even though schools exist for children yet children never drive cars to school.

The petrol station adjacent to the mini-roundabout adds more uncertainty. This station has a total of three entrances and exits, one in a side road and two of which divert traffic onto and off from the mini-roundabout. Passing drivers may also be distracted by reading the advertised fuel prices.

While both cyclists and pedestrians have no choice but to use this junction to get across the town, there is no provision here for cyclists to ride around the junction safely nor even to help pedestrians to cross the road.
The mini-roundabout is in the distance in this image, this is the road along which my mother was going to cycle in order to get home. Is this an equitable way to allocate space ? Both the mother with children and the person in a disability buggy have to use extremely narrow pavements while drivers enjoy wide lanes. The closest safe provision for crossing the road as a pedestrian is behind the camera.
Just like in other towns in the UK, planners here have only considered motor vehicles as a mode of transport. However, even that has not been done very well.

British Road Design
The roads in Burnham-on-Sea are very typical for a British town. I rode across this junction hundreds, if not thousands, of times when I lived in Burnham in the 1980s and 1990s. There has been no significant change to the roads in thirty years.

While British roads have not changed in design in the last thirty years, that is not true of the Netherlands. Sustainable safety principles applied progressively to Dutch roads in the ten year period between 1998 and 2007 are credited with having reduced the fatality rate by 5% per year. While the number of injuries for car occupants in the Netherlands is at around the same level as in Britain, pedestrians and cyclists are safer here than in the UK. Dutch children are far more likely to be independently mobile than British children and they are much safer than their British counterparts as they make their way around.

Comparison of the same road as above in Burnham (left) and a similar road and junction in Assen (right). The overall width of pavement, cycle-path and road at this point in Burnham is somewhat narrower than the example from Assen. Nevertheless, despite having less space overall, the British example allocates more space for cars even at this narrow point: the driving lanes in Burnham are about as wide as the driving lanes in Assen and the car parking lane, even though no parking is allowed. Extra width could be allocated to pedestrians and cyclists even here, but that isn't done even further along the road where it's much wider. Note also the smaller corner radii into side roads in Assen to reduce speeds and make crossing easier and lower speed limits in all residential streets. The width of the driving lanes of this street in Assen were reduced in 2008.
Burnham-on-Sea is a small town (the population is less than a third that of Assen at under 19000 people) but the roads in the town are always busy with cars. As in most of the UK, there is little alternative but to drive. Very few people will ever feel comfortable to cycle when the infrastructure design makes it so obvious that drivers are preferred over cyclists. In these conditions, people often believe driving to be a necessity.

When we visited in October, we saw that cycling provision is still virtually non-existent in Burnham. Enthusiastic cyclists ride their bikes, but there is little to attract those who are not enthusiastic cyclists. Many of the people who we came across who cycled used the pavement and apologised for doing so, or avoided the direct routes by using back streets and narrow pedestrian only cut-throughs.

The red line at the left shows part of the route taken by my mother to get from home to the town centre. The green line on the right shows the only cycling infrastructure in this area, a cycle-path built on the edge of a new housing development on the Eastern side of the town. This single path is of relatively good quality by British standards but it's not very useful because it skirts around the town and doesn't provide a direct route to common destinations. A very finely spaced grid of high quality cycle-paths is required to enable safe mass cycling. The blue line on the far left shows where Sustrans tell cyclists to ride along the beach.

Where the best cycle-path in Burnham
crosses a minor side-road, the corner
radius is large to enable high speed
when entering the residential streets
and drivers have priority. Also note
how much space is allocated for
driving vs. other modes.
A recently built housing development on the Eastern edge of the city provided a single cycle-path. However, this path is compromised in quality (narrow, gives way at all side-roads, low social safety) and doesn't actually go anywhere very useful. It skirts around the edge of the new homes where it was easy to build rather than going where it would provide a useful route.

There's a stark contrast between the housing development in Burnham-on-Sea which is typical for Britain, and what we've seen with a new development built on the edge of Assen which is typical for the Netherlands. In the Dutch example, cycling is not an afterthought. Not only was this development designed to enable most children to cycle to school but the new suburb also came with all facilities including a pedestrian and cycle friendly shopping centre. What's more, not only does the new development in Assen have an extensive network of internal cycle-paths, but it its construction also involved extensive works to provide a direct, unobstructed and nearly car free route to the city centre by bicycle.

Is everything perfect in the Netherlands ?
The collision in Assen took place at
this junction in an industrial estate
We go to some lengths to point out, both on this blog and on the study tours, that the Netherlands is not perfect. While the overall standard of the Netherlands is very good, there are plenty of places in this country where the infrastructure is not yet up to scratch. Plenty of work remains to be done. We have observed that those places which have inadequate, old-fashioned, infrastructure are often the scenes of collisions.

This is a long, straight road with lanes
just as wide as those used in the UK.
Drivers often exceed the speed limit
of 50 km/h in this location. No cycle-
paths here to keep cyclists safe.
This was illustrated in the last week because a friend of ours was involved in a very similar crash here in Assen to that which befell my mother in Burnham-on-Sea. Like my mother, our friend was bruised but not seriously injured. Her bicycle had some damage. The car driver in this case initially wanted to leave the scene of the accident but did co-operate.

It's often the case that people from outside the Netherlands overestimate the effect of "Strict liability" law. This was not a deliberate act and law does nothing to prevent accidental crashes. The driver simply "didn't see" the cyclist when he used a junction to turn around. No change to the law can prevent crashes because they address the wrong issue. The law does provide the useful service of ensuring that the crash victim will have her bicycle repair paid by the insurance of the driver.

Proper cycling infrastructure in this location almost certainly would have prevented the crash from happening because this removes the chance for a collision to occur. There is easily enough space for a cycle-path along this route and it is highly likely that one will be built in the future. The Western end of this area is to be rebuilt in the next few years and I expect to see changes for cyclists when that happens.

The route taken by the Assen cyclist through the industrial estate is shown as the bottom line, in red. The location of the crash is shown by the blue cross. No-one used this as a through route until very recently when a new bicycle bridge joined the short green section of cycle-path on the right to the red road. There are many cycle-paths in Assen and they go to all destinations. The longer green line across the top shows one good quality route (mainly cycle-path) to the same destination. This shortest safe route is 0.8 km longer for this particular journey, hence the desire to save time by choosing the red line. As the red route now makes a useful bicycle through-route, it is imperative that a cycle-path is built alongside it to improve safety.
Bad road designs are unsafe
Further along the same road, still in the
industrial area, there's a painted mini-
roundabout (new since we moved here)
Dutch drivers cut corners just as British
drivers do. This type of infrastructure
is no more safe here than in the UK.
Bad infrastructure in the Netherlands is no more safe than is bad infrastructure in the UK. Where conflict is caused by road design, problems of safety also arise.

In the post second world war period, the Netherlands and the UK were transformed by the rise of the private car. Starting in the 1970s, the Netherlands took a different path. Some forty years have now elapsed of this second phase of transformation. People now take priority and road safety has improved by a huge margin.

In some places, old fashioned and unsafe infrastructure remains in the Netherlands. In the case of the street featured above, it affected relatively few people until the recent opening of a cycle-bridge made it into a short-cut for some journeys by bike. It is now important that this street too needs to be updated to make it safe - part of an on-going process.

Book a Study Tour in order to find out about
what works and what does not work for cycling
in the Netherlands.
Just because something exists in the Netherlands, that doesn't mean it works well for cycling. Many things can be found on the road network of the Netherlands or are promoted by Dutch companies but that doesn't mean it's a good idea to copy all of them. This is why we offer independent advice about infrastructure. We're in a unique position to understand and explain, having experience of the UK and the Netherlands, and we have no company's product to sell.

My mother and Kona are not the only cyclists in Burnham-on-Sea. This town is also home to the world's oldest triathlete.  On the theme of the last paragraph above, some may remember that I criticized a Dutch company last year who offered "Dutch" infrastructure which was not the same as they would build in the Netherlands. Public companies are amoral. Just because they have Dutch roots that does not mean they're interested in cycling. Profit is their motive. The same company has now become involved with achieving consent for a shiny new nuclear power station just 10 km from my mother's home. That's not very "Dutch" either - there are no nuclear power stations under construction in the Netherlands near the mothers of the people at the top of Royal Haskoning.


Unknown said...

Dare I admit I work for one of the other Dutch civil engineering companies - I believe its name means something like 'muck shifting'.

Anyway, the problem we have in the UK as designers is we need some sort of evidence or standards to base our designs on. You could say it's self evident that the Dutch have cycle infrastructure correct - use the CROW manual - unfortunately unless agreed otherwise with the Client, that can only be guidance.

Therefore, in the absence of standards, reference to UK relevant research might seem a good defence of an engineering decision. However, what quality of research do we have in the UK? I don't think it's jingoistic to say TRL used to be the pre-eminent research body in the World when it came to transport, but if you look at their latest study of cycle infrastructure PPR 580 'Infrastructure and Cyclist Safety', from 2010, it would be laughable if its conclusions and the consequences weren't so serious. It seems unable to distinguish the differing levels of segregation and talks about the fact that segregated facilities are more dangerous than on-carriageway cycling, referring to studies which looked at combined, unsegregated, footway cycleways. It also concludes that segregation is dangerous because at cycleway crossings of side roads in Finland motorists don't give way to cyclists as the law says they should! That this is down to law breaking by motorists (and probably poor design of the crossings) rather than some intrinsic failure of segregation doesn't merit further attention. It's just more evidence that segregation is bad.

It actually gets worse since, having used apples to prove oranges are dangerous they state "...the evidence regarding segregated facilities suggests that they are likely to prove attractive to some cyclists but could result in a net increase in risk to cyclists unless the speed of traffic is controlled at the points where they meet the highway network". They're almost there, they almost get
that it is traffic and the interface between motor vehicles and cyclists that's the problem...then they shy away from that conclusion. Motor traffic is the elephant in the room.

However, finally they acknowledge in 'Infrastructure as a System' that the likes of the Dutch have much lower levels of risk for cyclists...but...having come to an erroneous conclusion previously, they are at a loss to now explain this, since after all segregation is dangerous, they already know that, they've stated so in this very report. Their exact words "The emphasis [on the reduction in risk] in these studies on segregated facilities [in the Netherlands and Sweden] at a system-wide level seems to contradict the equivocal evidence regarding individual segregated facilities...". Except of course the 'equivocal' evidence they used either doesn't relate to what the Dutch provide (nor which anyone else recognises as segregation) or they seem to have focussed on the wrong danger - not that presented by motor vehicles and their interface with segregated cyclists.

If I didn't still hold TRL in some regard I'd be tempted to think the conclusions came before a word of this report had been written.


bz2 said...

Andrew: Many Dutch councils do recognise that in some circumstances (in particular, at roundabouts with separate cycle infrastructure) the risks of separate facilities are possibly higher than without, but they build them anyway as they feel the increased perception of safety and the extra convenience of not having to yield will attract more cyclists, which outweighs the downside of the increased risk of (low-speed) collisions.

I hope this comment makes it through, Blogger had some... issues earlier.

David Hembrow said...

Andrew: thanks for your comment. One of the greatest problems we face is a lack of "quality control" by people who write such documents.

It's simply not possible to find any sensible result from combining bad results which result from bad design with good results which result from good design. It flies in the face of logic to attempt to do this, yet we see it again and again.

This is why I've tried to distinguish _good_ Dutch infrastructure from the less good both on the blog and on the study tours. On the tours we go to some effort to explain that the bad examples which we demonstrate are not to be copied. The idea is to show people the types of problems which they can avoid.

bz2: While both decisions about roundabouts result in conditions which are safe relative to other nations, the debate about roundabout priority in the Netherlands is interesting. In Assen, cyclists lose priority at every roundabout (this is not as bad as it sounds) while in Groningen, cyclists have priority at every roundabout. As a result we get to see and use both kinds regularly.

Actually, in Groningen there's one particularly bad example of a roundabout. The issues there go well beyond priority and we spend some time there on the study tours, watching how and why this confuses motorists and cyclists alike and fails to keep people safe.

Overall, the Netherlands is still doing very well and remains the single best country in which to find good examples. However, I think it's important that we recognize the challenges which remain.

Unknown said...

It's sad cyclists have to mess around places like where your mom had the accident - a roundabout connected to the gas station and no bike lanes, that's utterly terrible. I can't imagine more stressful environment for myself as a cyclist.

I agree totally that separating the bikes from the general roads is the best way suitable. First of all, as a cyclist, I must say that involving me into roundabouts feels awful, it causes me a lot of stress usually. Imagine here in Toronto we have some bike lanes that go centre of the street (end of the article). Can't really believe how that should allow me to drive more safely - maybe I'm all wrong, but the idea of rear view mirrors rolling both sides of my handlebars doesn't make me feel exactly safe.

I really think that a cyclist should himself be choosing roads that avoid regular traffic and, on the other hand, the municipalities should try to make sure there are such ways to go.

Ride safely :)

David Hembrow said...

Michal / Martin,

Canada unfortunately doesn't seem to have yet come to grips with what is required to make cycling safe and pleasant. Last year, I wrote a critique of the many problems apparent from even just a short read through the new Ontario Bicycle Facilities manual. You may find it interesting.

Lanes in the middle of the road are simply a mad idea. You can't access anything from the middle of the road without having to cross traffic. It's not convenient at all. The Netherlands used to have some lanes like this, decades ago, but they're gone now. It's not something to be copied.

As for choosing less busy roads, this is something which is not always made easy. Cyclists need the most direct route. That's why when cycle paths are not appropriate, the Netherlands unravels driving routes from cycling routes so that cyclists can have the shorter option.

Unknown said...

Yes, you're right, it feels a lot like Canadian municipalities are positive/green/etc, but usually have not much idea of what to do, which results into something like you describe :-/.

On the choose-your-way I'd like to use Vienna, Austria, a very cycle-friendly city, as an example. They have a regulation that allows cyclists to use one-way streets freely in any direction, which seems really useful: 1) you always see the car approaching you 2) the small streets have the cars driving along slowly so even if you managed not to notice, you should be safe in case of contact 3) riding through quiet roads you feel less nervous and can actually ride faster, so even if taking a bit longer route, you might cross it pretty quickly. This is really a million-dollar idea, any bigger city should learn from this!

David Hembrow said...

Michal, please don't use Vienna as a target to aim for. It's not actually even good for Austria. Vienna is a city with a mere 5% modal share for cycling which in a country which also has a 5% modal share.

Conditions in Vienna are unpleasant enough that cyclists from that city often complain about them in comments under this blog and others. There is also at least one blog from Vienna which illustrates the problems. Watch some of these videos to see some of the problems.

If you want to progress you need to aim higher than this. While this blog post criticises one isolated place which is bad here in Assen, the majority of this country is vastly in advance of what you will find in Austria or any other country. The Netherlands still provides very easily the best examples to copy.

As for one-way streets,being able to go the opposite way along then by bike is not new in the Netherlands. Indeed, that's the case for all one-way streets here in Assen. Note, though, that's it's not done simply by saying that cyclists are allowed but by first restricting the volume and speed of motorised traffic on those same streets. As a result, the situation for cyclists on those streets is rather better than for Vienna.

That is the "million-dollar idea", not merely putting up a few signs.