Friday, 5 August 2016

Speed Bumps are not effective "traffic calming" but an ineffective "Band-Aid" for poorly designed streets. Examples of more effective ways to calm traffic

A few days ago I tweeted about how speed bumps are used in the UK as a band-aid on streets which should have no through traffic. This perhaps requires a little more explanation.

No-one likes speed bumps
Speed bumps are unpopular everywhere. There are claims that they delay emergency services, even that they can injure some people transported by ambulance. Speed bumps cause damage to cars and lead to drivers slowing down to cross the bump before accelerating, which increases exhaust emissions and noise pollution. What's more, speed bumps are very unpleasant to cycle over and can cause danger to cyclists.

Most speed bumps shouldn't exist
Sadly, speed bumps seem to have become synonymous with traffic calming in many peoples' eyes. The problem with installing speed bumps as traffic calming is that the "solution" doesn't match the problem. Rather than attacking the cause of traffic problems, especially on residential streets where problems are typically due to their being used by through traffic, traffic is allowed to continue while an attempt is made to regulate speed by means of speed bumps. This approach has very serious limitations.

Learn more about how the Netherlands made
mass cycling possible. Book a study tour.
Assen barely uses speed bumps at all
The city where we now live provides many good examples of cycling infrastructure, road and residential street design. We operate cycling infrastructure study tours from this location in order to demonstrate these concepts. In comparison with a British city of a similar size, Assen has very few speed bumps. See the photos below for examples both of poor and good practice

Bad examples
First the bad examples. Standard disclaimer: These photos and a video demonstrate where "traffic calming" is ineffective. Don't copy anything from them. Click here to go straight to the good examples.
An example from the UK. In an attempt to reduce speeds, seven speed bumps have been installed in a street of less than 500 m. However the problem with this street is the through traffic and these speed bumps do not address that problem. Using this street allows motorists to avoid a longer journey with traffic lights and is a rat-run especially during rush hour.
Another example of bad practice in the UK. This street has six speed bumps in 500 m but again becomes a rat-run during rush hour because it accommodates through traffic. The problem is exacerbated by allowing bidirectional usage when the street is so narrow. I once had to push my daughter off her bike between parked cars along here because a driver heading towards us at excess speed clearly had no intention of slowing.
Speed bumps are sometimes combined with restrictions (horizontal traffic calming) such as that shown here. But these also do not address the problems caused by allowing through traffic. Note that this street in London has both a 20 mph speed limit and is posted as a through cycle-route. This doesn't mean that conditions are good for residents or for cyclists as you can see from the video below.

Residents of Gillespie Road in London are campaigning for better conditions, using video of the rat-running traffic in their street. No-one should have to live with behaviour like this on their doorstep. I wish them luck with their campaign.

Speed bumps are not only used in obvious residential streets in the UK, but also sometimes on through roads like this. The road looks like a race-track but the speed bump requires slowing to around 10 mph. This bump rather took me by surprise on my last visit to the UK when I drove here in the dark at the posted speed limit of 30 mph. This is again bad design. There is no reason why drivers are expected to slow sharply at this point. Locals will learn about this. Otherwise it is unpredictable.
Good examples
Now for some good examples, all of which are residential streets in Assen which don't have a traffic problem. Assen has almost no rat-running. In the case of newly built areas, opportunities to take short-cuts though residential streets were designed out. In the case of older areas they have been re-designed out. Streets are "calmed" by means of reducing traffic and by designing streets which do not encourage high speeds, not by causing motorists to have to abruptly slow for obstacles.

If you would like to copy this success, rather than risk losing the context of these examples "in translation", we suggest a study tour in which you can be shown how these examples and others work.

30 km/h residential street in a 1970s suburb. While this is at times a busy through route by bicycle, there is no through route here for motor vehicles so motor traffic is restricted to residents and visitors.
50 km/h road in a 1970s suburb. There is more traffic here than on the 30 km/h street, but the traffic here also is residents and visitors only. There are no through routes by motor vehicle anywhere in this suburb and therefore there is no through traffic. Note raised table for the pedestrian and cycle-crossing
1970s suburb, brand new asphalt. The surface is suitable for high speeds, but the road layout is not. This is a non-through route. A cul-de-sac with no traffic other than that of the neighbours. Children play safely on the street here.

On the left, an example of a street in a newly built suburb. The yellow street on the left has been designed to encourage low speeds both by choice of surface material and by eliminating straight lines. At the same time, the white cycle-bridge makes cycling more attractive by providing cyclists and pedestrians with more direct journeys than are possible by car.

Retro-fitted junction between residential streets in a 1950s suburban area. The entire junction is a raised table. Note bollards used to encourage drivers not to cut the corner.
The reason why streets in a 1950s suburb do not suffer from rat-running is that the majority of the streets have been made into an elaborate one-way system specifically to break up through routes. The intention was not to make cycling difficult and therefore cyclists are always excepted from the one-way system. All the one-way streets in Assen serve to reduce motor traffic and they are therefore not comparable with those in the UK and other countries.
Another example of a one-way street in an older residential area in Assen. The one-way system is used to eliminate through traffic, except by bicycle.

Smoothly raised "roundabout" junction between residential streets in an older area. This makes it obvious to drivers that they should not speed, while causing little if any problem for cyclists.

The Netherlands has the most extensive network of low speed streets in the world. Almost all residential streets in Assen (and across the Netherlands) have 30 km/h speed limits. However, it's important to note that the calm nature of these streets is due to the removal of traffic more than it is due to the speed limit. Merely posting a lower speed limit is not effective.
This residential street is close to the city centre and perhaps attractive to people trying to avoid parking fees. Parking bays are provided for residents. The car on the left is not legally parked. Read more about residential car parking in the Netherlands. This is one of Assen's rare speed bumps. It's smooth to cycle over through not ideal with three wheels.

As we get very close to the city centre, where drivers would find parking on residential streets to be desirable, parking is for residents only (zone 6). The blue sign with a red bar bans on-street parking from that side of the road. The allowed side of the road for parking alternates, and this emulates the planned meandering road as seen in the newer suburbs.
In 2009, this obvious through route in a residential area had an obvious speed bump with cycle-bypasses. In itself, this is reasonably well designed. However, it was not necessary because this street does not actually suffer from much through motor traffic. See the next photo.
By 2012 the bollards had been removed and the speed bump had been made smoother. In reality, while this is a main route for cyclists (it continues as a bicycle-path from the point ahead with the circular blue sign) there is very little motor traffic here. Through traffic is removed by the one-way streets elsewhere. Note that while this street is one of the rare places in Assen where cyclists "share" the street with buses there is no bus-stop bypass.

People often mistakenly believe that the tiled surfaces of Dutch roads are historical. Actually, they're often very modern. This photo shows a street being transformed from wide asphalt to narrower tiled surface with off-street car parking. Read more about residential car parking.
A street which was transformed slightly earlier than the example in the last photo. Note that this is part of a one-way system which excludes through traffic.

In an older residential area, one of the narrowest streets in Assen. This is again part of a one-way system which excludes through motor traffic. By excluding through motor traffic, conditions can be made good for walking and cycling even on the narrowest of streets. Note how parking is allowed on one side of the street only. This alternates, emulating the planned meandering road as seen in the newer suburbs.
We've demonstrated these ideas to study tour participants for many years. In this example I'm explaining about one-way streets in Assen on a study tour. On the next Study Tour (September) we will again demonstrate a variety of residential street treatments.
Not everything is perfect in the Netherlands
While the Netherlands leads the world in such things as calming of residential streets, that does not imply that everything here has equal value. The Netherlands is a work in progress and problems can be found here too. These two examples are of residential streets in Assen in which there are problems, though these problems are at a much lower level that those illustrated by the British examples above:
A plastic doll, Victor Veilig, is requesting that drivers using this street should slow down. That a resident felt the need to buy this doll and install him on the street is an indication that all is not well.
Planters used as horizontal traffic calming. This street is not massively attractive to through traffic because it parallels a much more suitable road. However, these planters would not have been installed unless they were believed to be necessary and that indicates not all is well with this street. 
Even the relatively smooth and well designed speed bumps in Assen cause problems for some users. Any bicycle with more than two wheels, such as this machine being used by a person with a disability person, can't fit around the side of the speed bump. That problem is magnified when there is also an illegally parked car.


zmau said...

Though I perfectly agree that one-way-systems design is preferred solution, I could not say speed bumps are that bad.
Speed bumps - if designed well - do not make any problem to cyclists. It is clearly visible in some of your examples, and you can see my examples here :
Also, when talking about increase in noise and pollution, they are much better than traffic lights (which are occasionally used for traffic calming), and they do not slow down pedestrians.
Raised pedestrian crosses from your examples are also a kind of speed bumps.
And bumps are clearly much cheaper (specially from land-use aspect) then meandering roads.

zmau said...

Please take a look at this example.
Street is definitely through route for car-traffic (it leads outside the city), and there's a school on the right-hand side. Parents of school children want to put traffic lights on crossroads, in order to protect children. My NGO suggests speed bumps as more elegant solution. What is your preferred solution for this issue, if speed bumps are just a symptom that something else is done wrong ? (Ok, surely it's wrong to put school and a busy road side by side, but that's what we have.)

David Hembrow said...

zmau: Unfortunately I have to disagree with you quite strongly. The design of speed bump which you prefer is amongst the most vicious for cyclists. Please take note of the last photo in my blog post above. How would this disabled cyclist use your street ?

I think you've misunderstood what I meant about one-way streets. The idea is not to have any cars using these streets at all. That's definitely better in terms of local pollution than any speed bump. If motorists find that they cannot use streets as through routes then they won't use them. Dutch one-way street networks are used in precisely this manner, redirecting cars away from those streets so that they do not go along them at all unless they are accessing a residence.

Meandering roads are easier to retrofit than you may have realised. All you need to do is to alternate the side of the road on which parking is allowed. This required nothing more than a few signs. Build outs can also be used. See some of the residential car parking examples.

David Hembrow said...

zmau: Replying to your second message, that's an enormously wide road and clearly a through route. Traffic may eventually be able to be redirected elsewhere, but that may be some time in the future. What needs to happen now is to narrow the road lanes and add cycle-paths. Look at the layout of Groningerstraat in Assen, which would appear to be narrower than your road. There's a school along here too, and thousands of children cycle here every day. It's also made safe by traffic light junctions of an exceptionally safe design.

André said...

Looking at the British and Dutch examples you provide, a clear difference is the place where the speed bumps are placed. The British examples have a number of speed bumps intended to slow down traffic on all of a specific street. The Dutch examples on the other hand are at specific places in the street, where there is either a crossing with another road, or a place where cyclists and/or pedestrians tend to cross, telling drivers that a slow down is necessary at that specific point.

zmau said...

Yes, it is problem for all 3-wheelers. And it can be partly fixed by widening gap between speedbump and curb to - say - 1 meter. I did not insist on such solution just because there are really few 3-wheelers in my town.
But if solution is excellent for 90+% of cyclists, I can say it's not bad.

"The idea is not to have any cars using these streets at all. "
As I understand the matter, there are always local cars there. And I guess they can make problems, too.
What is usual car-traffic frequency on such streets in Assen ? (Sometimes when I think about making such a system in my hometown, I conclude it will make paths through such zones longer for local drivers, so it can worsen things up, if not committed carefully.)

I like the trick with "alternating the side of the road on which parking is allowed" very much, and I am gonna try to use it, thanks.

David Hembrow said...

Andre: That is precisely the point.

Zmau: You're looking at a way of making conditions slightly less awful for 90% of existing cyclists. What I'm describing is something which makes it possible for 90% of the population to be cyclists.

There are several differences between the behaviour of drivers on the streets which they live on and the behaviour of those who live elsewhere and who are just passing through a street, amongst them these:

1. People who live in a street have to look out for their own family and their neighbours. They don't see everyone else on the street as anonymous people, as do those who are passing through.

2. When people are at the end of their journeys tend not to be so stressed or in a hurry. Those people are where they want to be. On the other hand, those who are just passing through have no reason to want to be in this location at all. In fact, they probably already wish they were elsewhere. They may well have a long journey ahead of them when they already know they are running late. This causes them to want to hurry.

3. If access is limited to those who live on a street then the numbers of vehicles will be very much lower than on a street which serves as a busy through route.

When there are traffic problems on residential streets these are almost always caused by drivers who actually have no reason to be on that street at all.

The "usual car-traffic frequency" on residential streets in Assen is approximately zero. That's how it should be. When through traffic is eliminated, you rarely see a car. Routes for cars are unravelled from both cycling routes and from residential streets. This unravelling is what civilizes residential streets.

Existing through routes can become far more civilized if the traffic is taken away from them and sent elsewhere. Let me show you an example in Assen (note that in this example the new road is right alongside the old. This is useful for this example in that it makes it very obvious what has happened, but in most cases the new route is more remote than this).

Unless you are planning to remove through traffic from where people live, you will not civilize your streets. Adding traffic calming features such as speed bumps will do very little to help.

Anonymous said...

Re: usual car traffic frequency on Dutch residential streets.
It's easy to calculate, once you know the traffic is limited to residents.
There has long been a mandatory upper limit to the amount of car parking spaces per home which can be built by the council in new development. It's fluctuated a bit, but is mostly around 1.3 - 1.8 parking spot per house (for residents and visitors both, so if a house has its own parking spot in a garage that gets taken off the total). So a residential street with 40 houses will have about 60 parking spots. This means at most 60 cars will move yhrough that street during the rush hours, when everybody leaves and comes home. As these usually are spread out over more than one hour, and not everyone uses their car every day & at the same time, the maximum in one hour will be less than that. And the rest of the day there will only be sporadic car movements, from residents who aren't locked into the rush hours to do their travelling.
Maybe you have to add the residents of a few neighbouring streets, if your residential street's way to a larger neighbourhood distribution road leads through one or two other residential streets. But it still should be a very low amount in total.
If there are more that 500 traffic movements per hour (including cyclists), a separate cyclepath is mandatory.

John in NH said...

Often times in the US speed humps (raised tables, crossings etc.) are used for traffic calming because neighbors refuse to support cutting the through route off for themselves. They hate speeding an traffic but not so much as to inconvenience themselves. So vertical and horizontal deflection is used at a much greater expense than putting in a permeable diverter...Are there examples where a former auto through-route was converted to a permeable diverter or dead end? If so were the neighbors supportive? Why? (In some ways since more trips are by bike it would impact them less, when 95% of your trips are by car it would impact neighbors here in the US much more...maybe that is it...)

David Hembrow said...

Anonymous: You have it right about average frequency of movement. Non through streets have very little traffic. However, I have very strong reservations about the idea of limiting residential car parking so severely as you suggest. Why ? Because I've seen the devastating result of limiting car parking in British developments. Providing few car parking spaces doesn't in itself prevent people from owning or using cars, it simply makes those cars a problem. The result is streets on which the pavements (sidewalks) fill with parked cars, where cycling and walking is difficult, and disputes between neighbours. Dutch developments which provide ample, even excessive, space for cars actually work out as nicer places to live and nicer places to cycle.

John: It sounds as if you've had the same discussion as I had in the street which I lived in in the UK which was a notorious rat-run on which it wasn't safe to let our children play outdoors because of the many drivers who used it as a shortcut. My neighbours also valued being "able to drive out both ends of the street" higher than their own and their childrens' safety. This was despite a very effective example of a restriction existing just half a mile away which had greatly improved people's lives in that other street. I think the difference in NL is that people have all seen the advantage whereas in other countries they have not. Driving is actually quite easy in NL. It's just that cycling offers advantages.

Anonymous said...

I agree that limiting car parking too much causes unpleasant and sometimes dangerous situations, and I'm not advocating for it. But it has been the rule that town planners have had to abide by, for decades. When I bought my house 25 years ago and had questions about this, my town's planners explained it to me as follows.
The Ministerie VROM set the limits which town planners had to abide by, if they wanted permission to build x-many houses. These limits were intended to discourage car ownership, and based mostly on the needs of the big cities of the Randstad. But they were mandatory for all town planners, even in small country villages wheŕe car ownership was more of a necessity because of long distances and lack of local amenities. That Randstad-centric view caused problems in those developments, to the disgust of local planners who knew what would happen but weren't allowed to solve it pre-emptively, or even after the problems became obvious.
I don't know if the rules still apply, but for most of the areas built between (at least - I don't know about earlier) 1980 and now it provides a good rule-of-thumb to determine how many cars can live in a street.
One way to get around the maximum in country places where building ground was cheap, was to add a garage and count that as 1 parking space per house, knowing that the residents would have enough room in their driveways/front gardens to park a second car but not counting that. That left the extra .5 places per house to realise some parking places for visitors.
But in places where ground prices were driven up by speculators, and they still wanted to build affordable houses for starters, the extra garage/driveway option wasn't feasible and it led to some troublesome parking situations.
I think it's still in effect, it was only a few years ago when the new developments in my town were built - those had land prices driven up by speculators and have parking troubles now.

Anonymous said...

Re: one way residential streets.
I'm all for them (well of course, I'm Dutch and used to them), and would very much like to see them adopted elsewhere.
Maybe it will help to convince planners in the USA, with their square street grids, that Barcelona has now reinvented this idea and called it Superblocks?
This looks more like the square street-grids they're used to, and they like things to be super! And Barcelona may have a better name with them than the (in their eyes) too liberal and/or provincial Dutch.
It also frames the improvements as not being for cyclists (I didn't see them mentioned specifically), but for livable streets for the residents and their children.
This is exactly what Mr.Hembrow has been explaining about the Dutch traffic calming with one-way residential streets and keeping through traffic on the arterial (distribution) streets. As long as they do allow cycling in both directions on the one way streets!
Even if they forget to put that in explicitly during phase one, if it turns out to be ubiquitous and not a bother (as it is here) it would be folly not to, when they get to the physical renovations phase 2.

It may be frustrating to have been advocating this for 10 years, and been ignored because cyclists are seen as a minority in England and the USA. But if reframing the advice as a solution for a livable city, not even mentioning cyclists explicitly at first, but just taking them into account regarding two-way permeability, gets the English-speaking world on board - wouldn't that be great?

David Hembrow said...

Hanneke / Anonymous: From what I can see, the advice to provide smaller numbers of car parking spaces appears to have been applied in the 1980s (e.g. in developments like this one), but not earlier in the 1970s (such as where we live) or in newly built areas, which actually more strongly resemble early 1970s layouts than those of the 1980s. Overly Randstad-centric planning will of course not really work for other areas, and actually I have to say that from the point of view of living in pleasant surroundings I'm very happy to live here rather than in the Randstad. Where is it that you live which has parking problems now ?

The Barcelona superblocks idea is good in many ways, but it's not really what I'm talking about. It seems to me that they've really re-invented a variation on the woonerf, which actually is out of fashion in the Netherlands these days. From what I can tell, no new woonerven have really been built in the Netherlands since the late 80s / early 90s.

A woonerf, a street on which children should be safe to play, should never be a through route. Not by car and not by bike either and if that's what they want to build then the idea in Barcelona of restricting speeds to 10 km/h does make sense. However, the result is that cycling on these roads will be inefficient so what they are actually doing is forcin cyclists to take the same detours as motorists, using all the same traffic lights.

Ideally, cyclists should not take the same routes as motorists and cyclists ideally shouldn't have to stop at traffic lights as if their routes are unravelled the traffic lights should only affect motor vehicles.

Anonymous said...

Thank you for your answers. Sorry for being anonymous, I don't have any if the necessary accounts. I live in Heerhugowaard. The newest developments there, Zuidwijk-Huygenhoek & Stad van de Zon, are cramped as regards parking, which causes problems in some streets.
I do know that land prices were raised a lot by land speculators, which made the builders cram in more houses to recoup their costs, and made it too costly for the council to keep more than the mandated minimum as public space. So maybe that's the reason, and not that old rule from the time my neighborhood was built.
Sorry for getting distracted off the topic of speedbumps.

I would like to thank you for all the work you've put into writing these articles and making your videos.
I discovered them a few months ago, and have been reading and watching the blog archive and video collection. It's been an eye-opener for me, so many things I never noticed or thought about. I find it incredibly interesting, and have taken to checking your blogs and those of Bicycledutch every week, as well as really looking at all the bike- and pedestrian infrastructure around me, and the planning decisions that are being made.
One of our traffic planners is a long-distance cyclist with one of those yellow torpedo-shaped recumbent bikes, so we can be fairly confident that any new plans will have adequate cycling provisions.

I hope you manage to get naturalised (is that the word?) quickly and easily, and thank you again for what you are doing with education about safe cycling infra.

David Hembrow said...

Hanneke: Glad to hear you like the blog. As you've probably noticed, I also have a "torpedo". As for naturalization, that's a long and fairly expensive process. I'm taking Dutch exams in a few weeks. Assuming I pass, I'll then be able to pay to ask if I may become Dutch.

I've only cycled through Heerhugowaard once. A few years back I took the train to Hoek van Holland rode back to Assen and stopped at Heerhugowaard on the way back. I specifically rode through Stad van de Zon because I knew it was planned to be CO2 neutral and I was particularly interested in seeing how that development looked. It's quite dense, and given the aim to be CO2 neutral it seems quite likely that they may well have looked to limit car parking here. But I thought the buildings surprisingly conventional in appearance next to other contemporaneous developments in NL.

Paul Cooke said...

the problem with British Speed Cushions is that you either have to be brave and actually ride over them in primary or even go towards the centre of the road gap to deter a stupid pass, especially before a pinch point etc.

If you go the the left hand side of the cushion, a pushy motorist will go for an overtake...

Here's one I particularly hate...

as usual, this through road is a nasty rat run of people avoiding the alternative of the ring road...

This road was an absolute delight for three weeks while they were repairing the weir further back. The entire road was access only to motor vehicles but permeable for cyclists and pedestrians...

sadl, it only took a couple of days after the roadworks to finish for drivers to realise their little rat run was open again...

troy said...

David, thank you for a great post. Inner-Melbourne councils adore speed humps, cushions and neckings. They like to 'encourage' slower speeds, and for people to use designated through routes. Of course, these facilities achieve neither. I live by a narrow, residential street that runs by a park with many children learning to ride bikes, and ends opposite a school. It's a crazy a PM rat-run. And yes, it's getting worse with connected way-finding apps that not only find, but publicize rat-runs. Speed humps are more appealing for those in a rush to exit than waiting at a traffic light. How could it be any other way? Your post has some great examples that I will forward on.