The Bow Roundabout is a busy and unpleasant junction in east London, with large volumes of traffic, including many bicycles and lorries. It forms a key part of the trunk network in London, distributing road-haulage around the city. River channels of the Lea Valley, industrial complexes, and the developing Olympic Park, block suitable alternative cycle routes.
Bicycles generally only want to travel between central London to the SW and Stratford to the NE. Lorries and other motorised traffic generally want to travel in all directions, but because of both a flyunder (NW to/from SE) and a flyover (SW to/from NE) they only need to use the roundabout if wanting to turn right or left. The exception is local bus services between central London and Stratford, which have a bus stop on the approach slip roads.
This fundamental difference in the routes between lorries and other motorists (who are always turning left or right) and cyclists (who are always going straight ahead) has likely been a factor that led to three cycling fatalities in the last year, at least two due to “left hooks” (lorries on a cyclist’s right turning left across them). The junction approaches have been remodelled several times, most recently introducing a traffic-light-controlled advanced stop box for cyclists, to ensure they get to the front of the queue to enter the roundabout. However, the design is non-standard and confusing for cyclists – tellingly there is an official video showing how to use it. The design results in a forest of traffic lights, and the “safe” advance stop box and subsequent cycle lane refuges are too often blocked with motor traffic, stopped by numerous further traffic lights, during busy periods. This video from The Guardian shows this to dramatic effect – skip to 3 minutes in.
The layout is also unpleasant for pedestrians, who have to cross two slip roads, but the first of the pair is uncontrolled – there is no traffic light for them. The junction is still not fit for purpose, but, crucially, redesigns cannot reduce the overall flow of traffic through the junction, as it is at capacity. Diamondgeezer has detailed the problem from a pedestrian’s viewpoint.
Here is the current layout:
Traffic lights are shown as black blobs. The pedestrian crossing points are showing with blue/white squares, and the pedestrian only section is shown as a dashed blue/orange line. To the east of the junction is a water channel, with a “floating towpath” beside it, shown with black dashes as a tunneled section. This route is not considered further here.
Here is my proposal:
Again, black dots are traffic lights, blue squares are crossing points, with an additional two blue squares showing bus stops, moved from the SW and NE slip roads. The road links to these bus stops are accessed only by buses and are shown in yellow. The cycle/pedestrian paths are shown as orange/blue dashed lines.
At first glance it somewhat looks like a roundabout in reverse, but this is not the case – there are four traffic-light-controlled “diamond” crossings, and traffic may only proceed straight across these.
This design improves on pedestrian and cyclist safety by properly segregating them from the road routes. For each direction, pedestrians/cyclists must make two crossings, both at right angles to traffic. At each crossing, there are two single-lane roads to cross, controlled by traffic lights which stop traffic on both roads at the same time, plus at the first crossing there is a bus road link which would be crossed via a raised zebra crossing, due to its low traffic levels.
Traffic negotiating the roundabout and turning left will meet just one traffic light, which would be green for just under 50% of the time. Traffic turning right will meet two traffic lights. Both will be red or both green, at the same time. This means that, again, nearly 50% of the time, traffic turning right would have green lights all the way through. There is no need for a dedicated pedestrian phase, with this design.
The only traffic which would not benefit from this design is the small amount of non-bus traffic travelling to/from the side-roads off the slip-roads to the NE, to/from the SW, and traffic travelling from the MacDonalds in the SW. This traffic would no longer be able to proceed straight across the roundabout to the NE unless it was allowed to use the small bus link-roads.
Disclaimer: I am (obviously) not a professional planner or traffic modeller, I am approaching this problem purely one of many regular cyclists on busy roads in London. Credits: The design was created in the Potlatch 2 editor on OpenStreetMap.org, and includes existing data from OpenStreetMap contributors and background aerial imagery from Bing. If trying this yourself be sure not to save your changes to the OpenStreetMap database. You can see the roundabout on an interactive map.
4 replies on “A Proposed Redesign for the Bow Roundabout”
Initially I thought from reading your opening lines you had read my ranting about this deadly design – even more deadly in its second iteration than the first, but you still have failed totally to actually go there and observe what cyclists are doing. Go and have breakfast at the Three Mills Cafe, sit outside and watch. then realise the total folly of going anywhere near that roundabout as a pedestrian or a cyclist
TfL’s own survey – referenced at the inquest into the death of Brian Dorling, noted that 60% of cycle traffic rides over the flyover. my own observations indicate this figure is on the low side – I’d reckon closer to 70% ride over the flyover. The flyover has a 4-lane dual carriageway but Westbound one lane is permanently closed off, and Eastbound just one lane was more than sufficient for the summer of 2012. Why is that? Well just bring up Bing Maps oblique view and Google Maps Street View and you’ll see that at least 80% if not more of the East-West flows of motor traffic are going to the roundabout. Aside from the buses going straight over 100% of the motor traffic circulating on the roundabout delivers the hazard of driving straight through the blue painted strip of CS2, at one of 4 slip road junctions, through one of 6 possible movements, where 100% of the cycle traffic going via the roundabout will be riding, and the only means of managing the risk of these two flows colliding is the assumption that both drivers and cyclist will act exactly as the traffic engineers have instructed them through traffic signals and road signage. There is no fail to safe condition if either makes an error of judgement.
I believe that TfL engineers actually put a proposal for sending all cycle traffic on CS2 over the flyover, where the hazard of having a motor vehicle drive across your path is – short of a suicidal idiot – is 0% of vehicle movements on the flyover. Go watch carefully, sometimes the ONLY vehicles in sight moving over the flyover are cycles. We have to seriously question why, who or what influenced the decision not to use this inherently safer option.
On the flyover the nearside lane in each direction can be taken over as a split footway/cycle lane vastly improving pedestrian access between Bow and Stratford. It isn’t as if it is that difficult to get across the slip roads- getting on at busy times the less confident cyclist simply walks over through the queue of stationary traffic, getting off at the East side there is a set of traffic signals regulating the flow. At the West side the signal controlled pedestrian crossing just the other side of the church also holds back the traffic flow -although ideally it could be synchronised with a second set of signals at the foot of the flyover ramp to stop the flow at the same time.
The solution is elegant in its simplicity and provides for cyclist riding with the main traffic flow to ride cleanly on and off the flyover, but the less confident, and pedestrians have marked, and possibly signal controlled crossing points, which make little difference to the traffic flow around the roundabout if they are synchronised with the signals at the roundabout feeding the slip road traffic on to the roundabout.
[…] […]
Love it! It seems like an excellent idea.
I note that the LCC’s “proposals to install safe crossing for cyclists and pedestrians at Bow were rejected because they would delay motor traffic too much.” It would be very interesting to see, therefore, what the transport modelling would show for this plan.
Before closing, I would like to leave you with a quote from Peter Wright, whose daughter Rosie was killed six years ago by a left-turning HGV:
“I think there is total incompatibility between trucks and cyclists. I think it’s impossible for cyclists to make eye contact with truck drivers–they’re not at the same level. Rosie was on the inside; there’s no way she can make eye contact with the driver of the truck. I don’t think he did anything other than he was just unaware that she was there: he couldn’t see her. I think basically, as in, for instance, Holland, cyclists need to be separated from that sort of traffic, not included in it and effectively have to make their own way through it, particularly in difficult areas such as that particular junction where most of the traffic is forced to turn left but she is able to go ahead.”
Thank you