Friday 30 August 2013

TfL's terrible "two stage right turn" - which is actually approved of by LCC

This blog post is primarily about a particularly poor suggested design for London. Click here for a particularly good design for turning across traffic, here for an example of a dangerous design,  here for a selection of good junction designs.

Yesterday evening, this week's Study Tour group, a mix of campaigners and planners from Australia, Britain, Canada and the USA, ate their evening meal with us at a nice restaurant in Groningen which happens to overlook the most dangerous junction in the whole of The Netherlands.

Why the shocked expression ? Read on
The food was great, the company was excellent, and while we watched how the traffic, especially bikes, crossed the most dangerous junction in the whole of The Netherlands, we saw several almost incidents which were enough to demonstrate why this junction is a problem. That's why we take people there. However, the shocked expression which Charlie Holland photographed on my face wasn't the result of what was going on in Groningen yesterday evening, it was the result of watching the latest video from Transport For London showing their plans for Stratford High Street, part of the extension of "Cycling Superhighway 2":

Is this the world's first cycle junction design to accommodate the needs of those people who are not "ambiturners" ? This junction design not only includes the appalling right turn design but also on-road cycle-lanes and advanced stop lines. These are also examples of old-fashioned and dangerous provision. Astonishingly, people actually get paid for this standard of work. If you're a Londoner it comes out of your taxes.

I've criticized two stage turns before because even if done well, they're hardly optimal to encourage cycling. Not only do they keep cyclists on the road, but they extend the amount of time that a cyclist spends in the junction area. Quite apart from any physical danger, this is certainly not good for subjective safety. A two stage turn is an admission of not knowing how to deal with cyclists. It results in delays, it's difficult with a cargo bike, when carrying children or towing a trailer to spin on the spot and enter the correct position to make the turn. If you're an older person or have a disability then the problems will be even greater. This design is particularly illogical, quite the reverse of the self-explaining streets which the Dutch strive for. It will not be easy for people, especially children, to understand so it will be used in ways the designer did not intend.

Some ASLs still exist in NL, but they're
amongst the things you should not copy
despite being "Dutch"
I've also criticized Advanced Stop Lines (bike boxes) before because these also cause conflict. A cyclist trying to position themselves correctly to make a right turn in such a box is required to ride across other lanes of traffic. The TfL designer actually expects this to happen as they've planned in an ASL which is in the middle of the road for right turns. This would not be required if they had designed the junction properly in the first place so that there was no advantage to cyclists from riding in front of buses.

"Being aware of each other" is not infrastructure, it's an accident waiting to happen. Crashes occur when people are not paying attention and it is the function of correctly designed infrastructure to be forgiving of mistakes, not to rely on people behaving perfectly in order to be safe.

Questions for the designer: Who is the target audience for this ? Who is it that requested their journeys be made longer and less convenient and who wanted to spend more of their time when cycling around junctions with motor vehicles ?

It won't necessarily be safe, but I expect
anyone running late for an appointment
will be tempted by the convenience of
the red line over the longer orange line
Do you really expect that someone will ride that extra distance to position themselves in your secondary waiting area if they're running late ? No-one should have to choose between "safe" and "convenient". This design forces people to have to make a "devil's bargain" when good infrastructure doesn't force that choice to be made.

Would you let a five year old ride across this junction ? Would you even let a teenager ride unaccompanied here ? If not, why not ? Is it not easy enough to understand ? Can errors easily be made ? These are admissions that the junction is not sustainably safe. Accidents will happen and injuries and deaths are likely to result from the building of infrastructure like this.

It's not just a sketch
According to the description on TfL's youtube channel, this is going to be reality in London as of October 2013. It is to be part of London's "superhighway" network.

Frankly, it's amazing that someone can do work of this quality, that they can pass it to their boss, it gets approval to be built, people can make promotional videos of it and the promotion department can let people know about it, and all of the people involved should have been paid for their efforts without anyone noticing how dreadful it is.

Transport For London, is this really the best you can do ? Do you honestly imagine that this will enable you to catch up with The Netherlands ?

What should they have done ?
The most advanced treatment of a traffic light junction in the Netherlands now is the Simultaneous Green junction. These junctions remove all conflict in both space and time. No-one has to find their way to the front of a queue of motor vehicles. No-one has to wait twice to make one turn. Everyone gets to take the most direct route across the junction, no matter what their skill level and without any risk of being run over. Children can ride safely across because it is obvious how to ride safely across such a junction. It's even possible to incorporate right turn on red for cyclists only and to make the phases of the lights such that cyclists' average delays are half that of motorists. There is a lot of inbuilt flexibility in Simultaneous Green junctions.
Yesterday, hours before dinner and hours before I'd seen TfL's video, our Study Tour group rode over this simultaneous green junction in Groningen. It's somewhat larger than the junction that TfL are planning for but extremely convenient and easy to use. Because no-one from TfL was in the Study Tour group, no-one from TfL saw this far superior and proven design of junction for themselves.
What's more, this design already exists and is proven at many different sizes of junction, from joining tiny roads (in Assen there's an example which has just one lane of motor traffic and a contra-flow cycle-lane) right up to junctions larger than that which TfL are dealing with, several lanes in each direction.

Simultaneous Green works for large
and small junctions. This road has
nothing but a short on-road cycle-lane
feeding into the junction and has on-
road car parking. It still works well.
In all cases, Simultaneous Green junctions are both convenient and safe for everyone to use. No-one is encouraged to choose a less safe route across the junction if they're in a hurry.

Before wasting money on doing the wrong thing, why didn't TfL send people over here to find out what real cycling infrastructure was, so that they were equipped to do a better job than they have ?

Not just London.
Southampton is also getting in on the act. More bad infrastructure design for a "superhighway", again making the mistake of separating "less confident" cyclists and giving them an inferior solution and again in a location which could very easily accommodate a Simultaneous Green junction which gave all cyclists both safety and convenience:
Complete with the three dangerous features of a bad bus stop designon-road cycle-lanes and advanced stop-lines as well as supporting "less confident" cyclists in a different and less convenient way than those who are "more confident", this junction design which Southampton claims is "Dutch" is actually absolutely nothing like real Dutch infrastructure. Why do British councils take tax-payers' money and use it to pay people who do such shoddy work as this ? Again, this would be far superior, far safer and more convenient for cyclists and also far more representative of real Dutch infrastructure if it had been designed as a simultaneous green junction.
Southampton, the design which you are considering is not "Dutch". You are being sold a copy of a type of junction which kills cyclists in Denmark. Please consider sending your planners on a Study Tour to find out about better solutions than this. Send them to find out what really is built in The Netherlands, how and why it works, so that you can apply the same principles in your city.

Click here to book a study tour
We demonstrate real Dutch cycling infrastructure over three days, showing some of the very best examples as well as demonstrating why the less good infrastructure should not be emulated.

Let us help you to avoid the costly mistake of building infrastructure which causes the conflict, danger and inconvenience which you have designed into the Itchen Bridge junction.

Saturday 31st August update
I've just found out that London Cycling Campaign actually approve of the ridiculous design shown at the top of this page for Stratford High Road. In fact, they're taking credit for it.  This is one of two things they describe as "solutions proposed by our Love London, Go Dutch campaign" (note that in response to this criticism they changed their web page at this link to say that it's a "poor implementation", but the headline remains as positive as before and comments under the blog post still refer to their enthusiasm).

Londoners, you've got to seriously think about this. You are represented by a campaigning organisation which sees "Love London, Go Dutch" as merely a slogan and nothing to do with actual Dutch infrastructure. Sadly, this has been obvious from the very start of their "campaign" when they admitted that one of their "Dutch" ideas was not Dutch at all and that they'd actually made up something they thought of as a "hybrid" of ideas from several countries but which actually doesn't appear in any other country.

I covered this a couple of weeks ago.
This one metre wide bus-stop bypass
with a post in it which is being praised
by LCC is not remotely the same as
"thousands in the Netherlands"
Another piece of infrastructure design which LCC are taking credit for is the inadequately designed bus stop bypass further along the same route (CS2). While LCC claim that there are "literally thousands of these in the Netherlands", that's simply not true. There are indeed thousands of bus stop bypasses over here, but they don't look like the one on CS2 - they're engineered to a far higher standard.

So long as the LCC continues to have such low aspirations for London, they will continue to be a part of the problem regarding cycling in the UK rather than being part of the solution.

I have invited people from London Cycling Campaign to join us on one of our Study Tours so that they can see and use real Dutch cycling infrastructure and have it explained to them. Thus far, they have declined the offer.

3rd October update
It's disappointing to see that instead of admitting that they've made an error, the LCC continue to try to spin this issue.

An LCC representative writes in a response to this blog post that "Well-designed two-stage right turns are approved of by both Dutch and Danish authorities and are common in Denmark". In fact, nothing remotely like the video from TfL above, which LCC says "could make cycling safer and more convenient" exists in The Netherlands. Remember that the campaigning position that LCC was forced by their membership to adopt was called "Love London, Go Dutch", not "Go Danish". This is not the first time that LCC has tried to confuse Dutch and Danish practice and their repetition of assertions like this one doesn't make them into facts.

It's not good enough for LCC to appropriate the language of change (as has happened with subjective safety and Go Dutch) but to carry on as usual, continuing to promote second-best solutions in London. While this continues, LCC remains part of the problem rather than part of the solution. These things need to turn into real policy and real change not just slogans.

I also note a comment made in the LCC response about not having been on our tours but going to "experts" instead. What LCC have actually turned down is the opportunity to take part in the only cycling study tours in this country run by people who speak English as a native language, who have lived and cycled in both the UK and the Netherlands and understand both countries and who are completely independent. We do not use the tours to sell the services of Dutch companies (note that employing a Dutch company is no guarantee that you'll get a Dutch quality of infrastructure). We demonstrate what does not work and why it does not work as well as showing those things which work very well. We don't give a false impression by visiting highlights but we aim to give an honest appraisal. We point out that not everything "Dutch" is worth copying, we don't rubber stamp anything and we won't be polite about bad ideas.

Frankly, I find the attitude of the LCC quite puzzling. If they were genuinely interested in furthering cycling, they would have investigated all possible sources of information, and that includes us. But they did not. Some people in LCC value our work enough to have asked us to do unpaid work for the organisation. They've also been happy to use our photographs and videos for free. But their interest extends only so far as what they can take. Apart from that, LCC has not made any contact at all.

We run our tours in order to try to educate people in organisations like LCC so that they can make better use of their funds and be more effective. Yes we do charge for the tours but this does not make us wealthy. In fact, we've barely done better than break even over the six years that we have operated study tours. That is as planned. This is not a self-serving exercise. Thankless as it often feels, this is something we invested a great deal of time and money into in order to try to make the world into a better place. When we've been able to afford to do so, we've even often offered free tours (i.e. at our cost) to people who thought would especially benefit.

If LCC staff are going to try to slight us, then it's only fair to point out that an independent source pointed out a couple of years ago that "70% of its budget goes on staff salaries and that the search for more funding is seen as an important LCC goal".

If anyone has their snout in the trough, I can assure readers that it's not us. We receive 0% of LCC's budget.

November 2013 update
In November, LCC returned to praising the CS2 extension of which this junction is a part, claiming it to be "a major success for LCC" before turning around and criticizing it a couple of days later after there had been fatalities on CS2. This is not good enough.

In case you're wondering, the most dangerous junction in the whole of the Netherlands is much less dangerous than several junctions in London. Nevertheless, this junction is an example of bad design and our study tours visit it because it is one of several places where we can demonstrate how conflicts are generated by bad design. TfL really would have benefited from this.


David Arditti said...

As the As Easy As Riding A Bike blog discussed, the TfL design is not really a two-stage turn, it is a four or five stage turn. I can't believe anyone will really do it as they hope. The worst aspect of it is the inbuilt conflict with pedestrians through the shared-space confusion. And you are right, the confusion is further accentuated by the provision of the advanced stop areas for (other?) cyclists in the main carriageway.

The Danish two-stage turn is, as you say, not as subjectively safe as the Dutch simultaneous green, or a Dutch protected two-stage turn. It is however a lot more subjectively safe than the normal UK method of making a right turn from the right-hand lane, and a lot more efficient than this TfL nonsense. But it, and the Dutch designs, only really make sense if there are at leat unobstructed cycle lanes, if not tracks, on all arms of the junctions. It is this lack, together with often having cars parked very close to the junction, that make these designs difficult to transfer to the UK.

David Hembrow said...

David, Simultaneous Green junctions are far more flexible than you think they are. Large and small junctions, wide and narrow roads.

I've added a photo above showing how the way into one of these junctions in Assen is nothing more than a narrow road with a very short cycle-lane at the junction and on-road car parking for the rest of the length of the road.

In another case in Assen the road is narrower. A single lane for cars on a one-way street with a contra-flow cycle-lane. This also works very well with the Simultaneous Green junction.

This would have worked very well at the junction which TfL wish to change. This is why I recommend it for that location. There is no need to look further for an inferior way of tackling this junction.

Unknown said...

I'd say to the designer, if you have to make an informational video to explain how to navigate it, then you're doing it wrong.

Jitensha Oni said...

So on the one hand we have "hardliners" like yourself arguing for rigorously Dutch infrastructure, and others arguing that this isn't practicable under current UK design principles and regulations, and we have to make do with what we've got. Unfortunately it is precisely the latter that is a major contributory factor to a) the glacial rate of progress in installing proper infrastructure in the UK and b) the mish-mash of styles, often unsafe, nearly always inconvenient for cyclists and pedestrians, and sometimes just bizarre, but always confusing, that is built to get round ths parlous situation. But under the current situation neither hardliners nor compromisers are going to get decent infrastructure in the UK.

So first things first. There is little point in anyone asking for more infrastructure or councils building anything until the correct design and regulatory framework is there. Otherwise you end up with massive inconsistencies. For example the transitions from the new Royal College Street to the surrounding bi-directional paths, neither of which are much good IMO, largely because they are only tuned to present capacity and demographic; the Walton-on-Thames town centre design which puts cyclists and pedestrians on a shared pavement which is busy with pedestrians most of the time, and of course this TfL solution which literally ties itself in knots. But to get around this one needs enough of a consensus that the safest design is that which is proven i.e. Dutch, or perhaps Danish. But until the UK government takes top-level ownership of the issue, I really can't see how the UK can have roads safe enough or coherent enough for mass cycling and walking.

aka Jitensha Oni

Paul Martin said...

It's quite astounding that so many locals believe the propaganda (and then think, erroneously, that the Dutch don't know what they're doing) instead of doing something which will allow them to see how it is REALLY done: a quick internet search!

Thanks for hammering away, David. We have similar problems in Australia and our geographical isolation makes things worse (why, I do not know...).

Richard Adamfi said...

Are simultaneous greens allowed in the UK without a change in the law?

Anonymous said...

I don't think there's any restriction in the Traffic Signs Regulations preventing a simultaneous green junction, but there might be a problem with the precise form of the cycle signals. This doesn't really matter, though, because DfT can (and does) authorise non-standard traffic signs on a per-site basis, so there's no need for a change in the law to allow a non-standard traffic light setup.

seiklmeikl said...

@ Richard: "Are simultaneous greens allowed in the UK without a change in the law?"

If so, what's the problem with changing laws?

London Cycling Campaign said...

Your headline is misleading because LCC doesn't approve of this design, which should be clear from our article which says:

"This is an overly complicated arrangement, which involves the cyclist mounting a pedestrian footpath, before entering an ASL in front of a stream of motor vehicles that could move at any time.

"While the cyclist has avoided having to cross streams of fast-moving traffic on Stratford High Street, other conflicts have been created."

"There’s also a strong possibility that cyclists will ignore the desired manoeuvre and instead move straight to the front of the ASL without looping around on the pavement."

Yes, we are pleased TfL is at last accepting that two-stage right turns are better than forcing cyclists to cross streams of fast-moving, but not the way these have been implemented.

I've adjusted the third paragraph to make our position crystal clear:

"Two-stage right turns and bus bypasses are solutions proposed by our Love London, Go Dutch campaign in 2012, and previously hardly seen in the capital, but these are inadequate implementations of those principles."

David Hembrow said...

LCC: I'm glad to hear that you've changed the text on your website but frankly, you're still far too complementary about something which is really not good enough. Neither of these designs for London are remotely good enough.

What's more, two-stage turns are not something that you should be championing at all. They prevent cyclists from making efficient journeys and rather than providing for everyone on a bike to make both safe and efficient journeys they encourage a choice to be made between "safe but not convenient" and "convenient but not safe". This devil's bargain type of infrastructure is a long way from what cycling campaigners should be calling for.

This type of infrastructure is something which we simply do not have in The Netherlands (I saw one more than ten years ago, but they were uncommon then and I've not seen one since) so why are you asking for this under a banner of "Go Dutch" ?

I've offered before to organise a study tour of real Dutch cycling infrastructure for the London Cycling Campaign. The offer is still open. Please come and see real Dutch infrastructure so that your campaign can be shaped by it. You're wasting a lot of time, money and effort in picking up and pushing inferior ideas like this one.

Nikephoros said...

Hi David,
Do you know if there is an actual forum there for "actual cyclists" in the English language? By this I mean for those who actual cycle for their day to day life needs, for those who are car-free or largely car-lite. In the USA where I live, everyone who touches a bicycle saddle for sport or recreation seems to consider (falsely) themselves a cyclist, but in reality if you use a car for all your necessary trips to work, shopping or school, how can you be anything but a motorist supporting the apartheid discrimination against all modalities of transport? Thus I feel most forums in the English language that purport to be for cyclists, are actually for motorists and their interests.

Sorry, for the off-topic post, but I figured you might have some knowledge on this.


Tim said...

Aside from the inconvenience it seems to me that the main problem with both these two-stage plans is that they are each one-of-a-kind non-standard rather complicated concepts (for those unfamiliar with the idea). As pointed out elsewhere, if you need a video to explain it, you've done it wrong.

In contrast, simultaneous all-directions green seems obvious to me. We've seen the equivalent in the UK for years for pedestrians, and even for a driver green essentially means "you can go where you like now" with the minor exception of a right turn (unless it has a filter arrow). Far more intuitive.

HB said...

Is it true that simultaneous green already exists for pedestrians in some places in the UK? And if so, wouldn't it just require adding bike signals?

It may be obvious, but how are pedestrians handled for Dutch simultaneous green intersections?

We do have a few simultaneous green even here in Los Angeles, with diagonal crossing, used by peds and bikes riding on the sidewalk, which is mostly legal here.

HannahC said...

Hi David - I think I'm aware of one example of a simultaneous green system in the UK - in Liverpool!

On my commute, there was a crossroads between a minor road and a busy major road. For the minor road, the council had put in a short kerb-segregated track leading up to the junction, with its own set of lights. The lights would feature in every sequence if you hit off the weight detector on the lane.

You can see it implemented here:,-2.956481&spn=0.006947,0.021136&t=m&z=16&layer=c&cbll=53.397563,-2.956445&panoid=y1s_CTp0D7iK9T0O4tMfpA&cbp=12,314.72,,0,-3.28

Now, it's not a perfect solution! The state of the road is terrible, and, astonishingly, there seems to be no provision at all for cyclists on the much busier main road (not even some paint!). I found that there was so little traffic on the minor road that the separate light sequence was unnecessary - the setup would have been much more useful for cyclists turning right off that main road! But it is at least a step in the right direction, no?

Unknown said...

There are often 'complaints' of cyclists going through red lights - indeed the police seem to think it a crime worth stopping - so the idea that cyclists will instead take a detour 2 or three times as long to cross a road seems a little silly.

Regarding the bus stop bypass, i'm sure better could be built, but there is a lack of space to build anything much wider.

I think the issue that we have in London is that Boris (or any politician) wants to do nothing that will piss of drivers - so what they try to do is to square the circle of fast roads with lots of cars, and safety for cyclists.

It's time to recognise we need less and slower cars on the road.

Multiparty Democracy Today said...

It is good that there is now segregation on the CS2 between Stratford and Bow Roundabout (the latter of which needs a massive update to include protected intersections. BTW, I am using protected intersection because the only turns you can make would lead you down a freeway), and soon to be segregation on the rest of CS2 to Aldgate, a massive improvement over what is there today, but why stop at what is being built or has been built on CS2? The junctions in particular still need to be improved, even over the two stage turns and separate signal stages and actually an early start that is not an always red for cyclists. Of course get rid of the latter, the early start, use protected intersections/simultaneous green, still with separate signal stages or at least turning traffic and cyclists if not full range of the intersection protected for bikes, widen the cycle tracks, use gateway designs at more intersections, use pedestrian crossing bypasses so as to allow bicycles to treat the puffin crossings as zebras, (and most likely needed would be a shortening of the red time for pedestrians), widen the bus stop bypasses, and probably widening the medians between cycle track and road. That they are using angled cubs is very good though.

Multiparty Democracy Today said...

I wonder whether the "approved and well designed two stage left (right) turns" referenced by the LCC is a reference to the protected intersection, which actually has a curb that creates a fairly well protected space to make such a two stage turn, or whether they really mean the kind that Copenhagen has. The protected intersection is genuinely used in the Netherlands on a regular basis at traffic lights which must remain, and is built today, but the kind that Copenhagen has is while used, is considered the granddad with the millions of euros in his will who just won't die and something that wouldn't be used for new installations.

Also, don't confuse cycle tracks are dangerous and bus stop bypasses are bad with poor implementation. The concept is what makes the Dutch able to cycle so well today. David tags these under "good examples", and Mark (Wagenbuur) also shows quite a lot of things that are good examples. You David and Mark argue over roundabout design for cycle paths and the protected intersection vs simultaneous green, but otherwise agree most of the time.