Friday, 16 February 2018

Alder Road, Poole: It's A Network Problem

Alder Road in Poole is being resurfaced and the local highway authority is making changes to the road marking layout to help people cycling. The problem is, it's not the road marking layout which is the problem.

Don't get me wrong. Making layout changes during major maintenance work is often a good idea, but unless this is following a grand plan, things get looked at in isolation and the best intentions can lead to failure. 

Alder Road is a typical urban UK road. It is a 30mph classified road (A3040) and as well as being part of a network of connector roads carrying some 18,000 vehicles per day (2016). It serves commercial areas, shopping parades and has extensive residential frontage. It also provide access to many residential areas and is a bus route.

The full 2016 Department for Transport traffic data is as follows (from manual counts, so accurate);

With 0.6% of the mode share, it is absolutely clear that those cycling are part of the fit and the brave group (or perhaps desperate) and I'm aware that the road is pretty gruelling as a commuter route.

In terms of geometry, much of the road has a 7.3m carriageway with two 1.8m footways - this can be found absolutely anywhere in the UK. I have had a very quick look at the casualty history for the road. The data isn't quite up to date, but the rate is something like 8 per year along the whole length with 1 being a serious injury and 7 being slight injuries per year. 

I've not crunched the numbers in any detail, but people walking, cycling and riding motorcycles feature highly in the data. In terms of traffic flow, people cycling and riding motorcycles are disproportionately represented given the low mode share compared to other traffic.

So, what is the issue with the resurfacing scheme? Well Alder Road traverses a series of hills and the idea is that where people cycling are going uphill, then they will be provided with a 1.5m advisory cycle lane to give them some "wobble room" (my language). This means that on those uphill sections, the general uphill traffic lane will be 2.8m and the downhill lane will be 3m. 

There are locations where the road width increases at junctions. In order to provide right turn lanes (turning pockets) and additional lanes approaching traffic signals, the advisory cycle lane stops. At signalised junctions, advanced stop lines for cycles are planned. As reported in the Bournemouth Echo, local cycling campaigner and trainer Jason Falconer has said;

"They just want it to look like they are doing something there... ...Large vehicles will have to drive in the cycle lane there, and the lanes will not continue over the junction where the road widens, it is really dangerous... ...It will put unskilled riders, the majority, at more risk than they are at present."

Jason tells me that he has made suggestions for this and other (better routes) which could be improved, but to no avail. Some of his thoughts are on his blog.

The local council, however, has responded by saying that a final decision hasn't been made. I have seen the plans (having described the approach above) and I have no doubt that the council has put in some effort to design a new road marking layout. However, the proposals are (in my view) doomed to fail.

Providing cycle lanes for travelling uphill can be of some benefit because people will be cycling more slowly and this means that they are more likely to move their position around (or wobble!). However, with a 2.8m uphill general traffic lane and opposing (downhill) 3m lane, the effect will be that those driving uphill will be pushed towards the centre line (which will no longer be in the geometric centre and because of oncoming drivers, they will want to keep further left. 

This means that at best, nearside wheels will end up wearing the advisory cycle lane and at worst, they will end up driving in the lane. A 7.3m carriageway is usually just split with a centre line giving 3.65m running lanes. The London Cycling Design Standards (4.4.2) gives some advice on lane widths and it turns out the 7.3m (3.65m lanes) carriageway is pretty poor for cycling;

The rule-of-thumb is to avoid situations where motorised vehicles and cyclists are expected to move together through a width between 3.2 metres and 4 metres.

For the Alder Road scheme, we at least have a total of 4.3m going uphill and 3m downhill which avoids the LCDS "rule of thumb" dimensions, but although those cycling downhill will be faster, the 3m running lane is going to feel pretty intimidating.

Anyway, we can debate the guidance, but the truth is that cycling has a 0.6% mode share on Alder Road. It is not surprising given the traffic volume and there is absolutely no way that the new road marking layout is going to improve that share. 

If we were going to change the road marking layout, I might be tempted to go with a less is more approach and only provide markings where they are really needed, such as at junctions, approaching traffic islands, bus stops, crossings and so on. Transport for London undertook a (limited) study on roads not entirely dissimilar to Alder Road which showed a statistically significant speed reduction. Perhaps Poole should save some money on paint and invest in some before and after speed surveys. 

From a cycling point of view, fewer road markings will mean that drivers might actually need to concentrate on giving space - if a centre line is present and there is oncoming traffic, some drivers still attempt to pass a cyclist while keeping to "their" side of the road. A lack of centre line requires more care.

I hesitate to make mention of installing lanes on both sides of the road without a centre line because on a road with these traffic volumes, it is a bad idea. I only mention and dismiss it because it is bound to come up in discussion!

So, I have suggested that Poole shouldn't proceed with their plan and indeed, I have suggested less is more, but what else can we do? Less is more won't make conditions worse for the 115 people cycling each day (or 58 people if they are going there and back!), but if we really want to grow cycling we have to look at how the network operates.

We don't want to convert the footways to shared-use, because apart from only being 1.8m wide, this would be a severe degradation for people on foot and it would be terrible for cycling. We therefore need to reallocate the road space. To provide a fully accessible and safe layout, we are going to have to make the road one-way for motor traffic.

This layout is extremely radical and would only be possible if the whole network is considered in terms of running local access through "traffic cells", re-routeing buses (probably with bus gates to maintain the traffic cells) and deciding that the very large A roads in and around the town will be the one taking through traffic. The arrangement will make loading a more troublesome and bus stops will need to use shared areas of cycle track as shown below from Royal College Street in London.

This is a very challenging proposition because we are so used to trying to bolt on cycling to either walking or motoring space. In situations like Alder Road, if we (rightly) accept we are not going to degrade walking space, then we have a stark choice to make with cycling. This something we cannot possibly do on a street by street basis and it is certainly not solvable with road markings.

My advice to Poole is that they should consider a "do minimum" scheme (with monitoring) for the short term and then think about what they are trying to achieve across the town. If they cannot begin to comprehend the scale of the challenge and the radical thinking needed, then they frankly may as well give up now. That goes for countless locations in the UK.


  1. In many ways the existing median with the broken white lines on both sides that exists for much of the road is already a hint to drivers that they should carefully consider passing. Do typical local drivers pull out over these medians or do they still try and keep to the left as much as possible?

    Median turning lanes are under appreciated for these low volume distributors, but thats if you want to increase vehicle speeds and throughput. Would tightening up the unsignalised T-junctions with bicycle bypasses be unthinkable?

  2. There is one other option that Germany uses which is to designate one pavement as a footway and wider pavement as a 3m wide cycleway/shared path with 'copenhagen'(continuous) crossings at side roads. This design is possible here but requires building out one footpath.

    Note it is not a shared path. It would be a two way cycle path and pedestrians would be expected to cross over to the other side at provided crossing points.

    1. I think that would be a tough sell in the UK, where walking is the main mode for short trips.

  3. 3m travel lanes, central hatching - bikes take the lane, cars overtake in the central hatching if no one is opposing (only recommended due to the low cycle volume)