CS2 from Bow to Stratford: Never Mind the Quality, Feel the Width (Part 1)

This was going to be one long piece about the new section of Cycle Superhighway 2 in east London, but I realised what a large task it is to cover all the weak points. So this is part one, you can read part two here.

I really want to be complimentary about TfL’s new flagship cycling facility on Stratford High Street, so I’ll start with the good stuff:

  1. The cycle tracks are generally of a good width*

So that’s the good stuff over with. I’m afraid to say that TfL’s highwaymen still have a lot of lessons to learn.

Never mind the accuracy, feel the length

Before I start banging on about all the juicy detail, I’d just like to point out that the route is around one mile long, and not the two miles which has been widely quoted by the BBC.

I don’t know where the idea originated that it’s two miles from Bow roundabout to Stratford – presumably a TfL press release has been stretching the truth by counting each side of the carriageway as a separate length.

It may sound pedantic, but Leeds and London are 200 miles apart. That does not mean the M1 motorway is 400 miles long. It’s 200 miles long. Each carriageway forms part of the same road. So the new bit of CS2 is one mile long, or around 1500 metres. (Credit goes to bikemapper for spotting that sleight of hand.)

The new CS2 is not what we have been campaigning for, is it?

So if the width is good, what’s bad? Well, nearly every detail of it is substandard, scoring from ‘less than ideal’ to ‘crap’.

Let’s start at the beginning.

Heading east from the deadly Bow roundabout (which I didn’t even attempt to use, by the way) the kerb-protected section doesn’t even start immediately. There’s a tiny bit of footpath-level cycle path round the edge of the roundabout but then it drops down to road level and runs for the first 300 metres or so as a mere painted cycle lane.

Here, this “Superhighway” crosses several side-roads and driveways, and there’s lots of construction traffic turning across the blue cycle lane. Large, fast, heavy vehicles pass within a few feet of you.

The footpath-level cycle path turns into a painted cycle lane. Note the HGV tyre marks have covered the blue within a day or two of it being painted.

The footpath-level cycle path turns into a painted cycle lane. Note where the HGV tyre marks have already covered over the blue at the junctions within a day or two of it being painted.

Soon we come up against our first bus stop, but as the segregating kerb hasn’t started yet (despite there being acres of space) it’s an old-fashioned one. The bus stop marking hadn’t been laid when I visited, but it was clear from the very narrow strip of blue paint that cyclists are expected to overtake stopped buses with no room for error whatsoever.

A photo of TfL's CS2 at Bow roundabout, with dangerous cycle lane and bus stop design.

Very disappointing. Imagine there’s another bus in this photo, stopped in the non-blue area on the left. That tiny gap is the “Superhighway.”

This would be a poor bus stop design even by 1990s standards. What the hell are TfL thinking installing this here, in 2013?

No cycle lane at all would be better than this, as the installed lane encourages riders to squeeze around stopped buses. And should a cyclist choose to ride further out, it will cause annoyance to drivers who are ignorant of why a cyclist isn’t using the “Superhighway” provided.

Dangerous junction design

For me, the biggest failure is the junctions. They all fall into the ‘crap’ category, many of them even into the ‘dangerous crap’ category.

It’s very important to get them right, but TfL have just resorted to painted cycle lanes rather than doing the job properly. Let’s have a look at one:

A dangerously designed junction on CS2, where there's no physical protection for people cycling

This isn’t what we’ve been campaigning for. Sure, cyclists have priority, but do you really trust London’s drivers not to left-hook here?

Anybody cycling along CS2 here needs the neck of an owl to pass this junction safely. Not only must they look behind to the right to make sure there’s no left-turning motor vehicles approaching from the rear, but a look forward-left is also required to check for vehicles emerging from the side road.

The segregating barrier finishes just below the bottom of the photo, so the only separation is white lines. Drivers turning left can then take a relaxed line towards the junction by drifting over into the cycle lane so they don’t have to slow down as much. Stratford High Street is a very fast road (and I saw no speed cameras) so some drivers will take this corner dangerously.

I expect that when driving along this fast road there is pressure on drivers to keep their speed up, so slowing down to take this corner safely – or stopping in the road to wait for cyclists to clear the junction – will not be a comfortable manoeuvre.

Why does the physical separation end so far back? Do TfL want to encourage dangerous corner-cutting by drivers here?

A far safer solution here would be to take the cycle path gently away from the road and for it to rise up to footpath height, so that anybody cycling across the side road here will be meeting turning cars at a right angle, which is much easier than the cars coming from behind. If motor vehicles have to mount a hump, and the corner radius is tight enough, then they will have to slow down and turn the corner at a much slower speed also.

This is more like it. Turning cars have somewhere to slow down and wait until the cycle path is clear, and the tighter corner radius and hump enforces slower speeds.

Something like this would be much safer. Turning cars have somewhere to wait until the cycle path is clear, and the tighter corner radius and hump enforces slower speeds.

Making Stratford High Street 30mph and installing some speed cameras to enforce it would also help matters, as drivers turning off the main road wouldn’t feel pressure to take the corner at speed to keep the carriageway clear.

Bust-up at the bus stop

Another poor detail is the bus stop bypasses. They are a compromised example of the concept.

On approaching the bus stop, the cycle path veers sharply away from the carriageway and narrows to as little as 1.5 metres. Due to the ridiculous decision to use 13cm-high vertical kerbs rather than the approximately 30º-angled, forgivingsplaykerbs, this narrow channel (or “gully” if you prefer) feels a little uncomfortable.

Also, due to TfL’s decision to keep the cycle path mainly at carriageway level rather than halfway between carriageway and footway level, the pedestrian crossing humps can be rather too sharp and steep, though this does vary from hump to hump. (There’s not much consistency along the route, except for the shade of Barclays-approved blue.)

Bus stop bypass on CS2 at Stratford

The cycle track width reduces from around 2m wide to 1.5m wide.

TfL have designed the bus stops this way to encourage lower speeds and force single-file cycling throughout them. I can understand why they’ve done this – there’s a lot of jerks who ride bikes in London, and we don’t want them hitting people crossing to or from the bus stop island.

But then there are a lot of jerks driving cars in London too, but TfL seem reluctant to apply the same logic to them, even though they can (and do) cause much more damage.

For some reason TfL are perfectly happy to let jerks drive two abreast at pedestrian crossings, but not cycle two abreast. I don’t understand why cycling gets stricter treatment than driving where there’s potential conflict with people on foot, but this isn’t the first time I’ve seen this (note the humps in the cycle track here, but complete absence of traffic calming on the parallel carriageway).

You may also notice the brown stripe along the blue paint in the photo above, which shows that the sharp bend doesn’t actually do the job of slowing cyclists down anyway – it’s possible to take the curve at speed.

*I say “generally of a good width” because it varies quite a lot along its length. I measured several points along the Stratford-bound side and the narrowest bit I found was a mere 1.3m where it narrows to accommodate a traffic light pole (though it is at footpath level so there are no kerbs, at least). The general width is usually at least 1.9m give or take a few centimetres and at the Stratford end it widens, bizarrely but luxuriously, to over 3m.


Coming up in part two (or three):

  • The “pants” junction and the loop-de-loop right turn
  • Signalised junctions with turning conflicts included
  • The bizarre ending
  • The ghost of the past
  • And other poorly planned cycling infrastructure

Don’t miss it!




Filed under Uncategorized

36 responses to “CS2 from Bow to Stratford: Never Mind the Quality, Feel the Width (Part 1)

  1. Not ridden CS2X yet and haven’t researched yet, but could some of the side road be no entry fro High Street to deal with left hook and access taken from a signaslised junction elsewhere?

  2. GareThugHowell

    I like your comments about the bus bypass and “jerks”. I can’t think of a better way of saying it myself. If you have got a Bicycle highway then it must be done properly; otherwise change the law regarding Bicycles as vehicles, ie so that they may interact with pedestrians as non-vehicles, and forget superhighways all together. The first two photos are bits of cycle “track” that I would not use. I would find some safe pavement, get off and push my bike. That’s what it is a push…. bike. Interaction with buses is always dodgy.

    Some of the cyclists who use these blue routes are going 30mph/
    13m/sec. How would that be against a pedestrian going 2m/sec? In an accident the cyclist continues to go forwards, ie over the handle bars, and does his best to protect himself. That would reduce his impact to, say, 5m/sec (13-8) The impact for the pedestrian would be at a velocity of 11m/sec, (13-2). I can see exactly why they want to slow the cyclists down by narrowing the track!

    I would not use the blue routes at least not at rush hours since there are too many speeding, and downright dangerous, cyclists.

    • If you don’t want to use the Superhighway then you’re in luck, as the footpath there is mixed use! And there are toucan crossings abound, too. I think they just threw everything at this area to see what sticks.

      I’m all for forcing vehicles (motor or otherwise) to slow down as they approach a potential conflict area but they’ve gone too far here for bikes at the bus stops (while not doing anything at all to slow motor vehicles where there’s potential conflict with bikes). Forcing single file riding is crazy.

      Due to the neglect of cycling infrastructure and promotion of vehicular cycling over the past decades, the majority of bike users in Britain are the fast and fearless who understandably don’t see the point in slowing down to pass a newfangled bus stop when they’ve just had to ride past dozens of the old-style bus stops in the traditional style.

    • Ghoti

      Narrowing cycle lanes doesn’t ony slow down bikes, it makes the bike lane unusable for trike riders (like me).

      • That’s a great point, and one which I hadn’t considered. Can you tell me how wide your trike is? I’d like to cover this point in the second or third part.

        • davidhembrow

          I don’t know how wide his trike is, but you may like to know that restrictions between posts etc. can never be closer than 1 m apart in NL due to the requirement that disabled people can use the infrastructure. We’d never see a width restriction at a bus-stop, of course, as that would be crazy.

          Very good blog post, btw. I “borrowed” your visualization of the junction (with credit and link). Hope you don’t mind.

          • Thanks David.

            Maybe the trike will fit through the bus stop bypasses without problems, though the sharp turns may cause difficulties. The 1.5m width of the track at the bus stops means single file cycling though, which is rubbish.

            You’re welcome to use the altered image wherever you see fit!

            • rich

              I’m a handcyclist and I can tell you exactly how well those sections work; Very poorly. If you’re on a shorter handcycle, like a Team Hybrid Cougar S24, or an Attitude, you can -just- about get around them usually if you get the angle right.

              If you’re on a full length handcycle, like a Force 3 – forget it, you don’t fit unless you are beyond precise and even then, you’re going to scrape regularly. This forces recumbents out into the road which is basically suicide.

  3. crapglasgow

    Can I just say, I think this blog is great. I’m glad some people have high standards. I don’t know why some people are praising this – those junctions are pretty much exactly the same as the ones that kill people on bikes up and down the country every other week. As you say, the Dutch best practice is (as usual) the only safe solution.

    • Thanks! I too am baffled by the applause this is getting. I suppose the basic concept is right – take away a lane of motor traffic to create safe space for cycling – but the details are all wrong. It looks better if you squint.

  4. T.Foxglove

    re: your design for the dangerous junction.
    #1 Great drawing. What software did you use for that?
    #2 I had a conversation with my local authority about a similar hostile junction (part of a winning bid of the Govt’s Safer Junctions funding in April) and was informed that they’d modelled traffic flow through the safe design & it only took a few cyclists in a stream to slow turning traffic proceeding, backing up cars up onto the main road and causing congestion. Also cars could take it at high speed & lose control on the speed table. My LA’s chosen design is similar to TfL’s, the intention being that a car will merge into the cycle lane prior to the turn & make the turn without holding traffic up. Sure, it’s a piece of infrastructure that is an accident waiting to happen but at least there won’t be a queue of traffic.

    • Jitensha Oni

      There will still be a queue if more than one car wishes to turn left and there are lots of cyclists. But what evidence do TfL or your LA have that these model scenarios would actually happen? A test with some temporary barriers (a planter or two?) in place of the paved area extesion shown by SC would soon reveal if there were an issue.

    • I used Photoshop, drawing lines and filling them in, over the photo on new layers (and copy-and-pasting the odd bit of road too).

      I expect TfL’s thinking was the same as your local authority’s. I can’t think of any other reason why the segregating kerb would end so far back. We’re still prioritising motor vehicle movement over human lives. It’s criminal.

    • George

      Merging into the cycle lane, assuming it’s a mandatory one, would be illegal.

      So that’s easily rebuffed as both dangerous AND inciting illegal driving!

  5. George

    My reaction initially was much the same.

    However, most of the junctions highlighted are relatively little used. This means for me as long as the markings are clear, the layout is okay. N not as good as a turn with a give way, but okay.

    The bus stop you mention with no bypass is not used or little used so I can see why they skimped, but they did skimp.

    The worst thing for me is the terrible engineering on the bus stop bypasses. The flooding is dire, the angles are dumb and the width slightly too narrow. Though I can accept some narrowing is a sensible cue to slow down a bit.

    I also don’t understand the lack of blue paint at some junctions. Blue paint is not very useful, but it’s more useful at junctions than on segregated sections.

    But this does feel safer than before-lots. Motorists seem to be behaving better at junctions. B+ from me.

    • “I also don’t understand the lack of blue paint at some junctions. Blue paint is not very useful, but it’s more useful at junctions than on segregated sections.”
      Exactly. Blue paint in the protected lanes is a waste. Where it is actually needed, i.e. at junctions, it is missing.
      One despairs at the sheer idiocy. How can an organisation be so dysfunctional that simple things like where to apply paint be done so wrong.

      • “However, most of the junctions highlighted are relatively little used. This means for me as long as the markings are clear, the layout is okay. Not as good as a turn with a give way, but okay.”
        I am sorry, but that is not a good point: it only takes an idiot to kill someone. Look at this: at another junction with a similar layout:

        An impatient inconsiderate driver.
        The truth is that British drivers are culturally not ready to “willingly” yield before turning left. Therefore, the infrastructure has to “teach” them who has priority. The mock-up does it very well; the actual design totally fails.

  6. Carla

    Thanks for posting!

  7. JZed

    Two big issues with the new lay out.

    First the segregation tries to force cyclists down to bow round about, trying to get across for the flyover is now significantly higher risk as you do not have the same distance to safely cross the lanes. Cars also now expect you to ride in the lane and not cross.

    Second, left hand turns. Before you could anticipate a left hand turn and move out to the right of the car, now the segregation forces you down the left of the car. For example the left turn at Warton Road. The flow of traffic turning left is high. I never road in the left hand lane, as inevitably the traffic is turning left. Now I’m forced up the inside of turning traffic who do not yield for cyclists.

  8. Pingback: Cycling on Main Roads – we need changes, not compromise. | nuttyxander.com

  9. “For me, the biggest failure is the junctions. They all fall into the ‘crap’ category, many of them even into the ‘dangerous crap’ category. It’s very important to get them right, but TfL have just resorted to painted cycle lanes rather than doing the job properly.”

    I heard Roger Geffen effectively endorsing this design on Wednesday, stating that the way the physical segregation ends early ‘reintroduces cyclists into the carriageway before the junction and gives them priority’.

    • davidhembrow

      And so the circle turns once again. While the most prominent campaigners actually praise bad infrastructure how can we expect standards to rise ?

      • Pete Owens

        Well the CTC used to take a more sceptical line on this sort of rubbish – unfortunately they have been swayed by the “go dutch” types in the LCC who insist that the only way to get people cycling is by installing segregated cycle paths with the inherent danger this causes at junctions.

        And whenever any of us sceptics point out these risks we are told that objective safety is irrelevant – what matters is subjective safety – ie the false sense of security due to non-cyclists mistakenly believing that this sort of thing protects them.

        • Come on Pete, surely you can see that a bad implementation does not disprove the concept. TfL have not installed any segregated cycling infrastructure at junctions, that’s the whole problem.

          The dangers at junctions shown in this post are not an inherent problem with cycle paths. They’re an inherent problem of badly-designed cycle paths.

          The only way to get people cycling is to install well-designed segregated cycle paths with well-designed junctions (along with traffic-restricted quiet roads, etc., etc.). There’s nothing else that has been proven to get millions of bums on saddles.

          Just because TfL can’t seem to get cycling infrastructure right does not mean it’s impossible.

          • Pete Owens

            The danger that cycle paths cause at junctions is very much an inevitable consequence of arranging for two streams of traffic to cross each others paths though a junction. In this case by placing a stream of cyclists heading straight on to the left of a stream of turning traffic. It is not a marginal problem you are looking at increasing the risk of crashing by a factor of 3. At least in this case you have with flow cycle paths; 2 way paths increase the risk by a factor of 10.

            The greater the degree of separation on the approach to a junction the greater risk as different road users are less likely to notice each other in time to avoid colliding.

            The problem has been known about for the best part of a century and no one has yet come up with an adequate solution. A number of measures to mitigate the danger have been attempted, and one of these is to merge the cycle path onto the carriageway on the approach to improve mutual awareness (though it should really be further back from the junction rather than the last minute affairs in Stratford).

            Of course at signal controlled junctions it is inexcusable that the cycle path is not given a stage in the cycle – so these could be made safe at the cost of delaying all traffic – including cyclists. However, the problem is intractable at priority junctions.

            • “The greater the degree of separation on the approach to a junction the greater risk as different road users are less likely to notice each other in time to avoid colliding.”

              You seem to be assuming that people will be speeding towards a junction. Well if you look at my altered images in part 1 and part 2 of this article, you’ll see that if you separate the cycle path from the road enough, you create two separate junctions. Make the corner radius tight enough so that cars have to slow right down, and they’ll meet any cyclists at slow speed and clearly visible as they’ll be at right-angles to each other.

              “The problem has been known about for the best part of a century and no one has yet come up with an adequate solution”

              I must have dreamt all that stuff in the Netherlands then.

              “merge the cycle path onto the carriageway on the approach to improve mutual awareness”

              You’re talking about my five year old niece merging with motor traffic at the approach to a junction?!

              I acknowledge that separate cycling infrastructure isn’t as simple than the UK’s carriageway/footway model. Of course adding that third category makes things more complicated. But it’s not so complicated that these problems can’t be figured out. Indeed, they have been figured out, and work wonderfully well, in the Netherlands.

              Surely the proof of the pudding is in the eating. Our vehicular cycling system has failed. We have few people using bikes (and those that do fit a very narrow demographic) and also a high KSI rate. The Netherlands has vast numbers of people using bikes, a very wide demographic, and yet the KSI rate is a fraction of ours. How do you explain that?

        • Pete, it is terribly sad that ‘the “go dutch” types in the LCC’ seem to have almost anything in mind other than Dutch quality of infrastructure.

          As Schrodinger’s Cat points out, none of the infrastructure built in London has properly kept cyclists apart from the danger of being on the road at junctions.

          Finally, you seem to have completely missed the point about subjective safety. Please read this explanation. No-one says that objective safety is irrelevant.

          The country with the best record on objective safety for cyclists just so happens to also be the same country as has the best subjective safety which encourages more people to cycle than anywhere else in the world.

    • Sigh. I wish Roger Geffen wasn’t such a nice guy, I’d find it so much easier to lambast outdated views such as that.

      Perhaps the old trick of replacing “cyclist” with ‘your child’ or ‘your grandmother’ works here: “The early end to the segregation reintroduces your child into the (fast dual-)carriageway before the junction…”

    • Pete Owens

      Stones and glass houses come to mind.

      Quiz Question:
      Who wrote on behalf of which organisation specifically in response to this dangerous crap design:
      “We are strongly in favour of the proposed cycle track along Stratford High Street, kerb-separated from motor traffic, and achieved by reallocation of carriageway space.”


      • Pete, I’m sure you are able to distinguish between principle and execution. The principle is good, the execution is bad.

      • As much as I support the work of the CEoGB, I had nothing to do with the article you linked to beyond letting them use my altered version of TfL’s image.

        As it stands, the cycle track itself is fine – it’s where it stops that there’s a problem. The article does mention the signalled junctions being poor, but they seem to have missed the poor unsignalled junctions.

        So I’m hardly a glass-house dweller chucking rocks about, am I?

  10. Pingback: CS2 from Bow to Stratford: Never Mind the Quality, Feel the Width (Part 2) | The Alternative Department for Transport

  11. Pingback: Dear John | The Alternative Department for Transport

  12. Pingback: Cycling on Main Roads - we need changes, not compromise. | Amble Scope

  13. Pingback: How to suppress bike riding #2: Bus stops (and also, the solution) | The Alternative Department for Transport

Leave a reply...

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s